Fisheries managers use stock assessments, or population models fit to demographic information about fisheries, to determine the health of the population and inform management. Management decisions are largely based on metrics called "reference points," which summarize the state of the population based on the fitted population model.
Reference points depend on population models primarily through an assumed stock-recruitment relationship, which describes how the population is thought to reproduce as a function of population size. Current stock assessment models often assume a two-parameter stock-recruitment relationship. But two-parameter relationships have been shown to limit reference points so that model estimates do not necessarily reflect the biology of the system. By adding more flexibility to stock-recruitment relationships, these limitations can be released.
In this project, Nicholas Grunloh is developing “metamodeling” techniques to understand the nature of how the assumption of common two-parameter stock-recruitment relationships bias reference point estimation (and therefore management decisions). By contextualizing current reference point estimates, this work highlights the need for developing more flexible stock-recruitment relationships.