Fin whale (Balaenoptera physalus) distribution modelling in the Nordic Seas & adjacent waters

Dec 8, 2019·
Diandra Duengen
,
Ahmed El-Gabbas
,
Elke Burkhardt
· 2 min read
Abstract

Understanding the dynamics of cetacean distribution in ecologically vulnerable regions is essential to interpret the impact of environmental changes on species ecology and ecosystem functioning. Species distribution models (SDMs) are helpful tools that link species occurrences to environmental variables in order to predict a species’ potential distribution. Studies on baleen whale distribution in polar regions are comparably rare, mainly due to financial and logistic constraints. Here we use SDMs to predict habitat suitability for fin whales (Balaenoptera physalus) in Arctic waters.

A combination of opportunistic and systematically collected visual observations from 2007 to 2018 was used. Opportunistic data were collected during ten RV Polarstern cruises in the Arctic Ocean (including the Barents-, Norwegian and Greenland Sea). Complementary visual data were obtained from open source databases. Environmental variables were chosen based on ecological relevance to the species, comprising both static and dynamic variables. We used MaxEnt software to model the distribution of fin whales, using presence-only data as a function of carefully chosen environmental covariates. MaxEnt’s predictive performance has been shown to be consistently competitive with the highest performing methods.

We were able to reveal important factors affecting the distribution of fin whales in the Arctic Ocean and how they respond to them. Results demonstrate the effective use of SDMs to predict species distributions in highly remote areas, constituting a cost-effective method for targeting future surveys and prioritizing the limited conservation resources. Results can be applied in a variety of purposes, such as designing marine protected areas and support the further use of opportunistic data to understand the ecological drivers of species distribution.

Type
Publication
World Marine Mammals Conference (WMMC), Barcelona, Spain. December 9-12, 2019
Understanding the dynamics of cetacean distribution in ecologically vulnerable regions is essential to interpret the impact of environmental changes on species ecology and ecosystem functioning. Species distribution models (SDMs) are helpful tools that link species occurrences to environmental variables in order to predict a species’ potential distribution. Studies on baleen whale distribution in polar regions are comparably rare, mainly due to financial and logistic constraints. Here we use SDMs to predict habitat suitability for fin whales (Balaenoptera physalus) in Arctic waters. A combination of opportunistic and systematically collected visual observations from 2007 to 2018 was used. Opportunistic data were collected during ten RV Polarstern cruises in the Arctic Ocean (including the Barents-, Norwegian and Greenland Sea). Complementary visual data were obtained from open source databases. Environmental variables were chosen based on ecological relevance to the species, comprising both static and dynamic variables. We used MaxEnt software to model the distribution of fin whales, using presence-only data as a function of carefully chosen environmental covariates. MaxEnt’s predictive performance has been shown to be consistently competitive with the highest performing methods. We were able to reveal important factors affecting the distribution of fin whales in the Arctic Ocean and how they respond to them. Results demonstrate the effective use of SDMs to predict species distributions in highly remote areas, constituting a cost-effective method for targeting future surveys and prioritizing the limited conservation resources. Results can be applied in a variety of purposes, such as designing marine protected areas and support the further use of opportunistic data to understand the ecological drivers of species distribution.
Maxent Species Distribution Models Conservation Mammals Marine Arctic Nordic Seas Barents Sea Fin Whale Sampling Bias