Estimating abundance of Risso's dolphins using a hierarchical Bayesian habitat model: A framework for monitoring stocks of animals inhabiting a dynamic ocean environment. (May 2020)

• Record Type: Journal Article
• Authors: Kanaji, Yu; Gerrodette, Tim
• Is Part Of: Deep sea research. Volume 175(2020)
• Publication Date: 2020-05
• Subjects: Grampus griseus -- Delphinids -- Distance sampling -- Species distribution model (SDM) -- Hierarchical modelling -- Western North Pacific -- Kuroshio; Oceanography -- Periodicals; Ocean bottom -- Periodicals; Marine biology -- Periodicals; 551.46
• DOI: 10.1016/j.dsr2.2019.104699
• Languages: English
• ISSNs: 0967-0645

Abstract

Abstract: Cetaceans often inhabit a dynamic ocean environment; therefore, a standard approach to estimating abundance faces the difficult task of discriminating between actual population trends and habitat shifts. In addition, because of the wide distribution of many cetaceans, it is often difficult to survey their entire habitat at one time, which makes it difficult to compare abundance estimates between years and to detect a trend. In this study, we used a hierarchical Bayesian habitat model to estimate the abundance and habitat distribution of Risso's dolphins in the western North Pacific. Based on information criteria, a model including depth, temperature, and yearly trends performed better than a model including only depth and temperature. The medians (with 95% credible intervals) of total abundance estimates in June were 54, 479 dolphins (25, 579–102, 086) in 2006, 54, 737 dolphins (26, 925–103, 932) in 2007, and 146, 179 dolphins (75, 352–264, 115) in 2014. Abundance was also estimated in August, but the seasonal difference between June and August was minor. Our models estimated that high densities of Risso's dolphins occurred in the mixed-water region off northern Japan and the cold-water mass off central Japan. These are productive waters because of their complex hydrographic features. Our results showed a probable increase in abundance and not simply a shift in habitat conditions or seasonal migration patterns. This study provides a framework for monitoring widely dispersed species in a dynamic environment, which can improve management and conservation in terms of more reliable trend estimation.