| Statistical Programs |
| College of Agriculture | University of Idaho |
| Seminar Announcement |
| "Applied Statistics in Agriculture" |
|
Hierarchical Modeling of Salmon Harvest and Ocean Migration
Presented By |
| Dr. Ken B. Newman |
|
Division of Statistics University of Idaho |
| Tuesday, October 14 3:30 P. M. Ag. Science 62 |
|
The development and evaluation of ocean fishery management plans
is facilitated by statistical models that link management actions
and biological processes. The essential data for developing models
are historical recoveries of marked salmon from samples of ocean
fishery catches. A hierarchical model, with the lowest level based
on a time series of mark recoveries by catch area, is formulated. At
the next level is an unobservable time series of abundance, distinguished
by spatial location. These two time series, observable catch recoveries
and unobservable abundances, can be described in parallel via a state-space
model. A stochastic model for the parameters which drive the harvest and
migration processes is the top level of the hierarchical structure.
In this talk I will discuss both the structural details of the model important to a fisheries biologist as well as details of parameter estimation of interest to statistical scientists. Key Words: state-space models, Kalman filter, empirical Bayes, spatial statistics, Markov chain Monte Carlo, importance sampling Postscript versions of related papers can be downloaded from http://www.uidaho.edu/~newman/ssm.html. All interested faculty, staff, and graduate students are invited to attend. |
| Return to Seminar Listing. |