Habitat suitability, movements, and diving behavior of the sei whale (balaenoptera borealis) off the coast of Brazil

Abstract

The sei whale (Balaenoptera borealis) is a large cetacean species that remains poorly understood, especially in tropical regions of the Southern Hemisphere, despite having been heavily exploited by the whaling industry during the 20th century. Listed as “Endangered” by the International Union for Conservation of Nature (IUCN), its population recovery and conservation depend on advancing knowledge about its spatial ecology, behavior, and habitat use. This doctoral thesis aims to address critical gaps in such knowledge by compiling novel data collected in the Santos Basin, southeastern Brazil, an area of high biodiversity but also subject to intense human activity. Species distribution models (SDMs) were applied to estimate habitat suitability for sei whales using true presence-absence data, and predictive maps were generated through ensemble models integrating machine learning algorithms. The models showed excellent performance, with mean depth emerging as the most important predictor, followed by sea surface temperature, surface current velocity, and chlorophyll concentration. Areas of highest suitability coincided with the continental shelf break, upper slope, and submarine canyon regions, emphasizing the role of static and dynamic environmental gradients in the species’ distribution. Patterns of occurrence, movement, and habitat use were investigated using satellite telemetry data from 26 tracked whales and analyzed through Hidden Markov Models, revealing that individuals alternate between two behavioral states: directional movements associated with transit and less directed movements linked to reproductive, feeding, or resting behaviors (Area-Restricted Search – ARS). ARS behavior was frequently recorded in topographically complex regions such as the Cananéia and São Sebastião submarine canyons and the Vitória-Trindade seamounts Chain. Dive behavior, recorded through satellite transmitters and high-resolution non-invasive sensors (DTAG and CATS), showed that 93% of dives occurred at depths ≤20 meters, within the draft range of large vessels, indicating a high risk of ship strikes. Generalized additive mixed models (GAMMs) revealed that dive depth and duration are modulated by both environmental variables (sea surface temperature, local bathymetry, chlorophyll concentration, diel period, and lunar phase) and behavioral factors (behavioral state and dive shape ). Occasional events of likely feeding near the surface suggest opportunistic foraging, even within areas traditionally regarded as breeding habitats. Altogether, the results of this thesis significantly advance the understanding of sei whale ecology in tropical regions, reinforce the importance of the Santos Basin as a recurrently used area, and raise concerns about the threats posed by intense maritime traffic and offshore industrial activities. Evidence of opportunistic feeding in breeding grounds challenges classical paradigms of functional habitat segregation and calls for more integrative conservation approaches. This thesis provides a robust scientific basis for marine spatial planning, mitigation of anthropogenic impacts, and identification of priority areas for the conservation of this endangered species in the Southwestern Atlantic.