Impacts of Misspecification of Spatial Structure of Assessment and Stock on Reliability of Reference Points
Principal Investigator:Tom Miller
Start/End Year:2017 to 2020
Institution:Chesapeake Biological Laboratory, University of Maryland Center for Environmental Science
Co-Principal investigator:Fellow: Reed Brodnik
Strategic focus area:Sustainable fisheries and aquaculture
The existence of spatial structure in populations of exploited marine fishes challenges our ability to develop reliable stock assessments. Using the northern stock of Black Sea Bass (BSB - Centropristis striata) on the US Atlantic coast as a model species, I will combine empirical and analytical approaches to explore the impacts of the spatial resolution of population and assessment models on the reference points generated by assessments. In this region, the distribution of BSB is highly structured during summer months when the fish are inshore, but the distribution is more widely dispersed when offshore in the winter. The extent and pattern of mixing during winter months is unknown, and thus challenges the ability clearly define the appropriate spatial structure for modeling purposes. In the work proposed for this fellowship, I seek to address three key objectives. Using otolith microchemistry I will compare the chemical composition of the otolith region associated with the previous summer period of fish collected at known locations offshore during winter, to the chemical composition of the outer margin of otoliths of fish collected the previous summer. This analysis will provide estimates of the mixing rate of BSB during the offshore period. Second, I will conduct a geostatistical analysis, with environmental covariates, of the winter istribution of BSB to quantify the vulnerability of fish to the offshore winter fishery, based on their pattern of summer residence. Finally, I will conduct extensive management strategy evaluation (MSE) simulations in a 2x2 factorial design to determine the impacts of incorrect specification of the spatial resolution of the assessment model in the presence and absence of spatial structure in the underlying population distribution on reference points used to manage the fishery. Working with Dr. Kiersten Curti at the NEFSC, the MSE will use an age-structured operating model with a six month seasonal time step, allowing the model to represent the spatially segregated summer distribution and the mixed winter distribution. Two scales of spatial resolution will be examined in the operating model: homogeneous, and a two sub-region inshore model with mixing during the offshore period. Data will be sampled from the operating model and subjected to either a single-region aggregated statistical catch at age assessment model, or an assessment model reflective of the two spatial regions. Assessments will be conducted from year 30 of 75 year simulations, and resultant allowable biological catches fed back to impact the operating model.
The results of the proposed work have implications beyond BSB, as spatial population structure is now recognized in several species along the US Atlantic coast. Moreover, improving our capacity to conduct spatially-explicit assessments is a goal of the National Marine Fisheries Service - Marine Fisheries Stock Assessment Improvement.