Modeling habitat use by Bryde’s whale Balaenoptera edeni off southeastern Brazil

Marine ecology progress series

, Sheila M. Simão¹  Rodrigo H. Tardin²  *, Yongwan Chun³ 

[1] Department of Ecology, Universidade do Estado do Rio de Janeiro, RJ 20550-013, Brazil;Graduate Program in Ecology and Evolution, Department of Ecology, Universidade do Estado do Rio de Janeiro, RJ 20550-013, Brazil. Laboratory of Cetacean Bioacoustics and Ecology, Department of Environmental Studies, Universidade Federal Rural do Rio de Janeiro, RJ 23897-000, Brazil. Department of Ecology, Universidade do Estado do Rio de Janeiro, RJ 20550-013, Brazil;Laboratory of Cetacean Bioacoustics and Ecology, Department of Environmental Studies, Universidade Federal Rural do Rio de Janeiro, RJ 23897-000, Brazil

关键词: Spatial autocorrelation;    Spatial eigenvector filtering;    Habitat use;    Distribution model;    Balaenopteridae;    Arraial do Cabo municipality;    Cetacean;   

DOI  :  10.3354/meps12228

学科分类：海洋学与技术

来源: Inter-Research

【 摘 要 】

Habitat-use models are a powerful tool for improving our understanding of the relationships between animals and their environment. With the development of GIS, these models have been used increasingly for the analysis of ecological data. However, they often suffer from inappropriate model specifications, particularly the assumption of independence, which is essential in conventional statistical models, and may often be violated during the collection of spatial data. Spatial autocorrelation occurs when the values of variables sampled close to each other are not independent, representing a major problem that must be accounted for systematically. We used a spatial eigenvector (SEV) generalized linear model framework to investigate the distribution of Balaenoptera edeni off Cabo Frio, in southeastern Brazil, an upwelling area impacted by human activities (tourism and fisheries). Sighting data were collected during 94 boat trips conducted between December 2010 and November 2014. A quasi-Poisson model using SEV indicated that the use of habitat by the whales varied with depth and the distance from the coast, and predicted that whales would be found most frequently around Cabo Frio Island and along the coastline, apparently overlapping with their prey. We found that habitat use was better predicted with the inclusion of SEV and that it is also possible to produce predictions of habitat use by correcting for spatial autocorrelation without the use of expensive surveys conducted by specialized research ships. This study provides useful insights into the habitat use of B. edeni in the southwestern Atlantic Ocean, and represents an important contribution to the conservation of this data-deficient species.

【 授权许可】

Unknown