Spatial and seasonal patterns of cetacean species richness: A Bayesian approach

Abstract

Research about community-wide patterns in cetaceans is challenging, so seasonal cycles and patterns of distribution are still poorly understood for many locations. In the present study, we proposed Bayesian Spatial General Linear Models (GLMs) to quantify and map cetacean species richness in southeastern Brazil's Santos Basin, considering environmental effects, effort, and seasonality. Systematic surveys were conducted in Santos Basin along zigzag-shaped transects over four years (November 2015 to April 2019), allowing consistent data collection over 40,000 km of effort through all seasons. A set of nine static physiographic data and eight dynamic environmental covariates were included in the models. A total of 371 sightings were recorded comprising 24 species. The best-fit model contained sea surface temperature, distance to 1000 m isobath, latitude, slope, and effort as covariates. We report that the cetacean species richness at Santos Basin is higher near the 1000 m isobath, in abrupt topographies, and off the northern portion; varies seasonally; is not spatially correlated; and is mainly affected by environmental predictors. These models and maps established a current characterization of cetacean species richness in these waters and have the potential to be applied in planning marine conservation, mainly in spatial prioritization programs.