Understanding the behavior and ecology of endangered species such as the North Atlantic right whale requires combining information from diverse monitoring efforts. Marine ecologists have collected heterogeneous data—such as aerial surveys that record whale locations and identifications, hydrophone networks that capture their vocalizations, and field samples measuring the abundance of their prey, zooplankton.

This seminar presents two complementary statistical frameworks illustrating how spatiotemporal modeling uncovers different aspects of whale ecology. The first employs a Hawkes process model to describe whale calling dynamics, distinguishing background communication from excitation responses and quantifying diel and noise-related effects. The second introduces a joint spatiotemporal species distribution model linking zooplankton availability with whale abundance through a hierarchical data fusion framework, enhancing inference on ecological interactions in a dynamic marine environment.

Together, these studies demonstrate how advanced statistical modeling transforms complex ecological data into coherent understanding of species behavior and ecosystem processes across space and time.