- This event has passed.
Advances in Statistical Modeling of Fisheries Data – Seminar
December 1, 2023 @ 2:00 pm - 3:00 pm
A Spatiotemporal Model of Condition to Derive a Novel Starvation Mortality Index That Improves a State-space Stock Assessment Model for Atlantic Cod on the Southern Grand Bank of Newfoundland
Madhura Weerasekera, Centre for Fisheries Ecosystems Research (CFER), Fisheries and Marine Institute of Memorial University of Newfoundland
The Atlantic cod stock on the Southern Grand Bank (SGB) of Newfoundland collapsed in the early 1990’s coincident with the collapse of cod stocks across Canadian waters. Consequently, a fishing moratorium has been in place for this stock since 1994. SGB cod is characterized by low productivity and poor fish condition relative to some other cod stocks. The primary focus of my research is to investigate how fish condition influences the natural mortality rate (M) and the prospects for rebuilding for the SGB cod stock. I developed a novel spatiotemporal condition model using DFO spring and fall survey measurements of fish length and weight. A stochastic approach was used to model weight as a function of length (i.e., condition) and other effects for survey strata, year, month, length, and their interactions. Model selection criteria were used to determine the best model. The model formulation that included only the interaction effects was the best fitting. This model was used to derive a novel length-based starvation mortality rate (MK) index. Results indicated that the MK index is higher in the spring than in the fall. It is also higher for cod between 55-80 cm in length and for cod larger than 120 cm. Additionally, the MK index was found to be higher during 1991-1993 when the stock experienced a substantial decline, and it was also high in 2016. The MK index was then converted into an age-based index and incorporated into an age-based state-space stock assessment model (SSAM) as a component of M. The inclusion of MK in the SSAM led to a significant decrease (62%) in the size of the cohort process errors (i.e., their standard deviation). This reduction suggests that the MK significantly improved the model of stock productivity. Therefore, I conclude that MK explains a substantial portion of the variation in M, and that starvation mortality is an important component of the productivity of SGB cod.
|Join from the meeting link
|Join by meeting number
|Meeting number (access code): 2772 604 8669
|Meeting password: mbGAJdJ8k73