This web application uses cookies and other tracking technologies to ensure you get the best experience.
SC/67B/NH/04
Download this file855 KB |
Download |
View directly in browser855 KB |
View in browser |
Resource ID
9453
Access
Open
Document Number
SC/67B/NH/04
Full Title
Preliminary report of abundance estimation for the North Pacific humpback whales using IWC-POWER data
Author
Kanako Inai, Koji Matsuoka and Toshihide Kitakado
Publisher
International Whaling Commission
Publication Year
2018
Abstract
The abundance of humpback whales in a summer feeding ground in the North Pacific was estimated using sighting data from 2010-2012 and 2017 surveys of the “International Whaling Commission Pacific Ocean Whale and Ecosystem Research (IWC-POWER)”. A design-based line transect method was primarily used for the estimation of density and abundance. In the estimation of detection function, both of half-normal and hazard-rate functions were used with covariates/factors such as “year of survey”, “school size”, “cue” and “visibility” as well as some likely interaction terms among them. The best model was then selected by an information criterion, AIC. The density and abundance were estimated by Horvitz-Thompson-like estimators to account for possible heterogeneity in detection process within the leg. In addition, a spatial modelling approach was tested as a model-based method using the GAM with potential covariates of longitude, latitude, SST and distance from the land to see any differences in estimates among two estimation methods. As a result, in the design-based model, the abundance of humpback whales migrating to the southern Aleutian archipelago (2010-2012 survey areas; north of 40°N, south of the Alaskan coast including both the US and Canadian EEZs between 170°E-135°W) in summer (July - August) was estimated as 9,868 (CV=0.53) under the hazard-rate model with the survey year, school size and visibility. The northern Aleutian archipelago (2017 survey areas; eastern Bering Sea) in summer (July - August) was estimated as 4,539 (CV=0.64) under the half-normal model with the school size and visibility. In total, 14,407 (CV=0.56). An initial trial for the model-based method provided an estimate in 2010-2012 survey areas as 16,961, which is somewhat larger than the estimate from the design-based estimates, and therefore it warrants further researches to improve the spatial modelling and compare the results. The authors conclude that the model-based estimates in the south area (2010-2012 survey areas) can be used for future in-depth assessment of this species in the North Pacific Ocean.