A Shark Habitat Model with Spatial and Temporal Dependencies

• Activity Number: 526
• Type: Contributed
• Date/Time: Wednesday, August 4, 2010 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Daniel Gladish*+ and Mark Wildhaber and Christopher Wikle and Scott Holan and John Froeschke and Gregory Stunz
• Companies: University of Missouri and U.S. Geological Survey and University of Missouri and University of Missouri and Harte Research Institute for Gulf of Mexico Studies and Harte Research Institute for Gulf of Mexico Studies
• Keywords: Ecology ; Bayesian ; Fisheries ; Spatial-Temporal ; Bull shark

Abstract

Statistical analysis on environmental habitat conditions have long been of interest for understanding relationships between environmental factors and species distribution. Complications arise when the species of interest resides in a marine environment, especially when spatial and temporal dependences are considered. The Bull shark is one such species that presents modeling challenges. The data were collected using gill-net surveys from 1975-2006 within the Gulf of Mexico along the Texas coast, with abundance of fish caught as well as physical characteristics of the water recorded. Previous work on this data used nonlinear approaches from a classical viewpoint. We present a data driven model allowing for spatial and temporal dependence with consideration for the relationship of certain physical habitat conditions and the distribution of the Bull shark from a Bayesian perspective.