Gedamke, T. and A. Seaver
The Survival Estimates In Non-Equilibrium situations (SEINE) model calculates mortality rates from changes in the mean lengths. The model is a variant of the Beverton and Holt (1956, 1957) annual mortality estimator which has received widespread use due to limited data requirements but has also been criticized due to the required assumption of equilibrium conditions (i.e. the mortality rate has been constant for enough time so that the observed mean length reflects the current mortality rate). Gedamke and Hoenig (2006) developed the SEINE model for application in non-equilibrium conditions and to allow the mortality rate to change at one or more points in time. Given von Bertalanffy parameters, the length at full vulnerability, and a series of annual observations of mean length over time, the model estimates mortality rates and the years in which they changed. A grid search over possible years of change is used to evaluate the likelihood surface and provide starting values for the final estimation. Additional changes in mortality can be added to the model and the improvement of model fit in relation to the additional parameters can be evaluated through AIC values. The model was developed for the assessment of goosefish (monkfish; Lophius americanus) where sample sizes for each year were quite low and the mean size was highly variable from year to year. It has since been used in similar data-poor situations including the assessment of the northeast skate complex and mutton snapper in Puerto Rico. The calculation engine was built using AD Model Builder by Alan Seaver and Dr. Todd Gedamke.