Using extended-duration audio recordings to survey avian species

Authors

  • Amy K. Tegeler,

    Corresponding author
    1. Department of Biology, Humboldt State University, Arcata, CA 95521, USA
    Current affiliation:
    1. Powdermill Avian Research Center, Carnegie Museum of Natural History, 1847 Route 381, Rector, PA 15677, USA.
    • Department of Biology, Humboldt State University, Arcata, CA 95521, USA
    Search for more papers by this author
  • Michael L. Morrison,

    1. Department of Wildlife and Fisheries Sciences, Texas A&M University, College Station, TX 77843, USA
    Search for more papers by this author
  • Joseph M. Szewczak

    1. Department of Biology, Humboldt State University, Arcata, CA 95521, USA
    Search for more papers by this author

  • Associate Editor: Koper

Abstract

Point-count surveys are widely used to infer avian presence and estimate species richness. Advancements in bioacoustic technology enable automated surveys that can supplement human-based point-count surveys with expanded temporal and spatial coverage. We surveyed birds in 13 Sierra Nevada and Cascade Range (CA, USA) montane meadows from May to August 2006 using 8 point-count surveys and automated audio recorders (ARU) to compare species richness between the 2 methods and evaluate the use of ARUs as a monitoring tool. We analyzed species richness using 30 minutes of ARU data per point and 2 point-count surveys. Automated audio-recorder data revealed 14 species per meadow (56 species total) while point counts detected 16 species per meadow (67 species total). Automated audio recorders provided >1,100 additional hours of data with personnel effort similar to 2 point-count surveys. An asymptote in species richness was reached for every meadow using ARU data and 8 of 13 meadows using 2 point-count surveys. We detected 81 species during all 8 point-count surveys. We used SonoBird (DNDesign, Arcata, CA) software to search for 24 species detected by point-count surveys but not in the manually sampled subset of audio files. We detected 22 additional species, bringing the total audio-file species detections to 85, 4 more than detected by 8 point-count surveys. We conclude that audio recordings and analysis provide an alternative to avian point-count surveys or as a supplement to increase their accuracy, particularly over larger temporal and spatial scales, or for species with low detectability. © 2012 The Wildlife Society.

Ancillary