Assessing bioacoustic techniques for inventory and monitoring of forest owls in the Central Sierra Nevada, California

• Main Author: Byrnes, Ryan M.
• Other Authors: Szewczak, Joseph M.
• Format: Master Thesis
• Language: English
• Published: California State Polytechnic University, Humboldt 2013
• Subjects: Occupancy; SonoBirdTM; Bioacoustic monitoring; Humboldt State University > Theses > Biology; Long-eared Owl; Great Horned Owl; Intelligent signal searching utility; Detection probability; Strix nebulosa
• Online Access: http://hdl.handle.net/2148/1647

Eight owl species of special concern occur in the Sierra Nevada: California Spotted Owl (Strix occidentalis occidentalis), Northern Spotted Owl (Strix occidentalis caurina), Long-eared Owl (Asio otus), Burrowing Owl (Athene cunicularia), Great Gray Owl (Strix nebulosa nebulosa), Flammulated Owl (Otus flammeolus), Elf Owl (Micrathene whitneyi), and Short-eared Owl (Asio flammeus). Additionally, four other owl species occur in the Sierra Nevada: Great Horned Owl (Bubo virginianus), Northern Pygmy Owl (Glaucidium gnoma), Northern Saw-whet Owl (Aegolius acadicus), and Western Screech Owl (Megascops kennicottii). Outside of the extensive demographic monitoring efforts focused on the California spotted owl and Great Gray Owl, there are no current inventory or monitoring efforts directed at the majority of owl species in the Sierra Nevada. In this study I explored passive bioacoustic methods and techniques for multi-species inventory and monitoring of the Long-eared Owl and Great Horned Owl in the Sierra Nevada. I focused on testing the methods and techniques within meadow vegetation types and surrounding conifer and mixed conifer-hardwood ecotones. During the 2009 field season I passively recorded owl species with Autonomous Recording Units (ARUs) at 50 sites divided evenly between Yosemite National Park and Stanislaus National Forest. Detection histories were constructed and used to estimate occupancy (the probability of a random site being occupied, ψ) and detectability (the probability of detecting an individual at an occupied site given the individual is detected, p) using program MARK and RMARK. Covariates such as meadow location (Yosemite National Park (YOSE) or Stanislaus National Forest (STNF)), survey duration, and meadow size were incorporated to model occupancy and detectability. Despite 50 sample sites and 11,781 hours of recordings, results suggest that passive bioacoustic methods may not be an efficient technique for multi-species monitoring of owl species given low detection rates for uncommon species ...