Joint Modeling Bat Assemblages Using Automatic Recording Devices

• Activity Number: 361 - Multivariate Ecological Data: Uncovering the Story in the Noise to Inform Species Management and Conservation
• Type: Invited
• Date/Time: Thursday, August 12, 2021 : 12:00 PM to 1:50 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Meridith L Bartley* and Kathryn M Irvine and Rogelio M. Rodriguez and Tomas J. Rodhouse and Benjamin D. Neece
• Companies: USGS and US Geological Survey and Oregon State University - Cascades and NPS - Upper Columbia Basin Network and Oregon State University - Cascades
• Keywords: Bats; acoustic monitoring; occupancy modeling; misclassification

Abstract

In order to understand and combat the conservation threats that North American bats are currently facing a collaborative monitoring effort was initiated in 2015 — the North American Bat monitoring program (NABat). This program relies on recording echolocation calls by remotely deploying acoustic recording devices. Automatic classification of bat echolocation recordings may include errors, leading to false negatives and false positive errors, however standard analysis approaches assume that misidentifications do not occur or the data processing workflow seeks to remove false positives. Current occupancy models that account for false positives allow for analyzing counts of call recordings from multiple species simultaneously while accounting for species classification errors. We apply this approach to an analysis of jointly modeled bat species surveyed in 2019 following a probabilistic NABat master sample. We account for unbalanced manual vetting across visits by incorporating an additional parameter for each species representing the probability that a call is capable of being manually confirmed.