Name: Samara Nehemiah
Date: 10/29/2024
Time (EST/EDT): 01:00 PM
Location: Chesapeake Biological Laboratory, BFL1101 (and zoom)
Access Link: email mees@umd.edu

Committee Chair: Michael Wilberg
Committee Members: David Secor, Geneviève Nesslage, Robert Latour, Amy
Schueller
Dean’s Representative: Jorge Holzer

Title: DEVELOPMENT AND EVALUATION OF SPATIALLY-EXPLICIT POPULATION MODELS FOR
ESTIMATING THE ABUNDANCE OF CHESAPEAKE BAY FISHES

Abstract: Although fish populations typically experience spatially varying
abundance and fishing mortality, stock assessments that inform management
decisions commonly model a population that is assumed to be well-mixed with
homogenous mortality rates. When assumptions about population mixing are not
met, these models can result in biased estimates. Spatial population estimates
are particularly beneficial to the Chesapeake Bay because this region faces
unique challenges as a result of climate change and fishing pressure. However,
use of spatial population models for fisheries management relies on models that
can reliably estimate biological parameters. Objectives for this research were
to 1) develop and implement a multi-stock, spatially-explicit population model
for Striped Bass (Morone saxatilis) to estimate abundance and fishing mortality
in the Chesapeake Bay and along the Atlantic coast; 2) assess the performance
of spatially-explicit models compared to spatially-implicit models (i.e.,
fleets-as-areas) to estimate abundance in the Chesapeake Bay, determine how
improved data quality (e.g., stock composition) affects model performance, and
determine the effect of aging error on model accuracy; and 3) determine how the
performance of spatial models are affected by potential changes in population
dynamics resulting from climate change (e.g., time-varying natural mortality).
The population model was a two-stock model with two sub-annual time-steps and
two spatial regions with stock and age-specific occupancy probabilities
representing movement into and out of the Chesapeake Bay. Fishing mortality was
estimated to be higher in the Ocean than the Chesapeake Bay and abundance
increased during 1982-2004 for both stocks before declining slightly until
2017. Simulations were conducted to test the ability of models to estimate
abundance and fishing mortality under alternative scenarios of data
availability and quality. Spatially-explicit estimates were approximately
unbiased when they closely matched the assumptions of the data generating
model. Models that ignored potential aging bias in datasets resulted in highly
biased estimates of abundance and fishing mortality. Although the performance
of all models degraded under most climate change scenarios, spatially-explicit
models produced the most accurate model estimates compared to fleets-as-areas
models. This research highlights the potential benefits of implementing
spatially-explicit population models for other ecologically valuable fish
species in the Chesapeake Bay.