Modeling audible detection of prairie grouse booming informs survey design. Issue 3 (19th December 2018)

• Record Type: Journal Article
• Authors: Holt, R. Douglas; Butler, Matthew J.
• Is Part Of: Journal of wildlife management. Volume 83:Issue 3(2019)
• Page Start: 638
• Page End: 645
• Publication Date: 2018-12-19
• Subjects: attenuation -- detection probability -- distance -- lek -- lesser prairie‐chicken -- monitoring -- sound intensity -- sound propagation -- weather -- wind speed; Wildlife management -- Periodicals; Zoology -- Periodicals; 333.954
• DOI: 10.1002/jwmg.21612
• Languages: English
• ISSNs: 0022-541X

Abstract

ABSTRACT: Many wildlife surveys, especially surveys of birds, are reliant upon audible detection of individuals. Many factors can affect call frequency and sound attenuation, which influence audible detection. Prairie grouse surveys are conducted during lekking and courtship (i.e., booming) to take advantage of frequent vocalization, resulting in greater detectability. Some researchers assume booming can be detected >1.6 km away. To test this assumption and determine factors affecting attenuation and detectability of booming, we established artificial lesser prairie‐chicken ( Tympanuchus pallidicinctus ) leks. We conducted sound detection trials at these leks using playback of recordings of male lesser prairie‐chickens calibrated to appropriate sound intensity. We modeled detectability using generalized linear mixed‐effects models with distance from lek and weather conditions as covariates. Our model suggested the odds of detecting lesser prairie‐chicken booming decreased 96.8% with each 1‐km increase in distance. Detectability decreased with increased wind speed and temperature, increased with relative humidity, and was influenced by wind direction. We use this model to predict probability of missing leks for line‐transects or point counts. For example, a survey of 1.6‐km‐wide transects during average weather (17.0°C and 60% relative humidity) with 12 km/hour winds would have a 25.6% (95% CI = 3.0–77.2%) to 77.6% (95% CI = 27.3–97.6%) chance, depending on wind direction, of detecting booming lesser prairie‐chickens within the surveyed strip. If point counts with a 0.8‐km radius were conducted under the same conditions, detectability would range from 17.7% (95% CI = 1.7–70.3%) to 70.3% (95% CI = 18.5–96.7%) depending on wind direction. To improve the overall chances of detecting lesser prairie‐chicken booming to ≥80%, surveys would have to be repeated 3 to 4 times. Our model describes the audibility of lesser prairie‐chicken booming and can be used to optimize tradeoffs among survey logistics (e.g., number of surveys needed, point spacing, wind speed thresholds) and detectability. © 2018 The Authors. Journal of Wildlife Management Published by Wiley Periodicals, Inc. on behalf of The Wildlife Society. Abstract : We modeled the audible detection of booming prairie grouse as a function of distance and weather conditions and our model suggested surveys conducted from 0.8‐km radius point counts would result in detectability <50%. To improve the overall chances of audibly detecting booming in an area to ≥80%, surveys would need to be conducted in wind speeds <12 km/hour and repeated 3 to 4 times.