Statistical Tools for Fitting Models of the Population Consequences of Acoustic Disturbance to Data from Marine Mammal Populations (PCAD Tools II)

Abstract

Abstract : Our goal is to build an ecological modeling framework that facilitates understanding of the way in which at-sea condition and health of various species of marine mammals changes over time. This project will develop statistical tools to allow mathematical models of the population consequences of acoustic disturbance to be fitted to data from marine mammal populations. We will work closely with Phase II of the ONR PCAD Working Group, and will provide statistical support to that group. Our scientific objectives are to build a statistical framework for understanding the way in which at-sea health varies over time for (initially) three species of marine mammals: southern and northern elephant seals, and northern right whales. Should the Working Group decide that we should address additional species, e.g. bottlenose dolphins, we will take those up in turn. For elephant seals our objective is to build and fit to data a hierarchical Bayesian model that provides daily estimates of lipid status. Maternal lipid status is a key variable in the life history of elephant seals, as it is strongly correlated with pup survival (McMahon et al. 2000). This model will use the drift dive behavior of elephant seals (Crocker et al. 1997) to provide observations on the underlying true, yet immeasurable, lipid state. For right whales, our objective is to build a model that provides spatially and temporally explicit estimates of individual health, movement, and survival. The model builds upon some of the ideas from the elephant seal project, but as the photo-identification of individual right whales is the core of the data, the model also includes many ideas concerning mark/recapture from (Clark et al. 2005).