Estimated density of Antarctic minke whales obtained from simulated IDCR/SOWER survey data using the “standard” method
Simulated survey datasets provided by Palka and Smith were analysed using standard line transect methods that assume g(0)=1. Estimates of mean school size were generally positively biased by >10% when cue production was asynchronous, but otherwise typically unbiased. Overall estimates of whale de...
| Main Author: | |
|---|---|
| Format: | Report |
| Language: | English |
| Published: |
University of Cape Town
2005
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/11427/19662 https://open.uct.ac.za/bitstream/11427/19662/1/Branch_Estimated_density_2005.pdf |
| Summary: | Simulated survey datasets provided by Palka and Smith were analysed using standard line transect methods that assume g(0)=1. Estimates of mean school size were generally positively biased by >10% when cue production was asynchronous, but otherwise typically unbiased. Overall estimates of whale density were negatively biased for nearly all scenarios considered, with mean bias of -23% for “2004” scenarios and -10% for “2005” scenarios. In the “2004” scenarios, bias was greatest when the detection function used to generate the simulated sightings included school size or weather as a covariate; in the “2005” scenarios, the greatest bias occurred when weather and whale density were correlated and when survey was in IO mode only. Implied values of g(0) ranged from 0.48 to 1.02 (mean 0.77) for the “2004” scenarios and from 0.78 to 1.03 (mean 0.90) for the “2005” scenarios, and were higher than previous estimates from the IDCR surveys (generally 0.5–0.7). |
|---|