Taiwan_offshore bat_2017Nov_2021Apr

Study Extent

This study was conducted at the candidate sites for the offshore wind turbine areas under the current renewable energy development plan of the Bureau of Energy, Ministry of Economy Affair of Taiwan (www.moeaboe.gov.tw/) in the Taiwan Strait off the northern and western coast of Taiwan Island . We concentrated our surveys in the area that was at least 500 m from the nearest land to avoid false detections of bats active at nearshore areas. The study area covered approximately 10,000 km2 with a maximum distance of 67.1 km to the nearest land

Sampling

Bats were primarily surveyed acoustically, using either a ship-based mobile transect line with a speed of 12-16 km/h or stationary recording from November 2017 to April 2021. Surveys were conducted with a Song Meter SM4BAT FS Bat Detector and a U1 microphone (Wildlife Acoustics, Inc.). The microphone was mounted on the top of the ship and pointed upward. The detector was deployed from 16:30 to 6:30 the following morning for each survey night, using trigger-recording mode before March 2018 and non-trigger recording with a duty cycle of either 1 minute per 3 minutes or continuously. We used a sampling rate of either 256 or 384 kHz with a high pass filter of 16 kHz. Direct observations that were recorded using a DSLR camera (Nikon D5) and a thermal imager (TiX 500 series, Fluke Coporation) in preliminary surveys in September 2017 are also included, to maximize our understanding of offshore bat diversity in the study area. All bat occurrence records were georeferenced by matching the timestamp embedded in the filenames or metadata with the tracking logs by an on-board GPS with time synchronized with the clock of the bat recorder and the cameras.

Method steps

1. Wave files were first split into short fragments with a maximum length of 15ms using Kaleidoscope Viewer version 5.1.8 (Wildlife Acoustics Inc., USA). Fragments with at least three, bat echolocation pulses, defined by the following settings: duration of 2~70ms, frequency range of 12~120kHz and maximum inter-syllable gap of 800ms, were used for further analyses. For each fragment, three to five search-phase calls with a signal-to-noise ratio of 10db in a single call sequence were used for species identification. Call morphology was inspected visually in the spectrogram with a setting of the FFT size of 512 points and 50% window overlap. In several cases, recordings were amplified using the Brightness adjustment (up to 15) for a better visual presentation in the spectrogram view, to maximize the descriptions of call features. Measurements of fifteen Spectro-temporal call characteristics, including two spectrogram features, four full spectrum parameters and 11 zero-crossing parameters, were taken from individual pulses of bat echolocation calls using the Analyze View function. Calls were assigned manually to the finest taxonomic level in Taiwan by comparison with reference calls of known bat species . Distribution range, global and local conservation statuses were collected from the IUCN Red List (https://www.iucnredlist.org/) and recent literature.

Seven bat species of three families

1. Pipistrellus abramus
   rank: species
2. Nyctalus plancyi
   rank: species
3. Eptesicus pachyomus horikawai
   rank: subspecies
4. Scotophilus kuhlii
   rank: species
5. Miniopterus fuliginosus
   rank: species
6. Tadarida insignis
   rank: species
7. Myotis
   rank: genus

Offshore to the northwestern Taiwan Island