A Surplus Production Model Database Incorporating Biomass, Catch and Environmental Data for Multiple Ecosystems: Construction, Projects, and Synthesis

Two international workshops have been held that focused on the production modeling approach, at various levels of aggregation and with various covariates.  How the outputs from such models could be used to estimate management-relevant metrics and ecosystem attributes was examined.  The objectives were to A) use comparative analyses to build a methodological foundation for evaluating how ecosystem structure and function interact to support fisheries production, B) determine what processes control spatial and temporal variation in ecosystem services, and C) use surplus production modeling as a unifying tool to evaluate the relative effects of physical forcing, trophodynamics, and fisheries exploitation on fish production.  To support these objectives, a database was created which incorporated biomass and landings data for 97 commercially and ecologically important fish and invertebrate species in 11 different Canadian, US, and Norwegian large marine ecosystems.   Climate data were also included and ranged from ocean basin drivers, such as the Pacific decadal oscillation (PDO), Atlantic multidecadal oscillation (AMO) and the North Atlantic oscillation (NAO), to regional environmental indices such as sea surface temperature and ice cover.  The database was refined at the workshop, and multiple tools were developed to aggregate the species in the database by habitat, size, feeding guilds, and common ancestry.   The database has facilitated a number of analyses including: A) comparisons of systemic, species-focused, and aggregate surplus production and indicators across ecosystems; B) empirical contrasts across levels of aggregation, drivers and ecosystems; C) the respective roles of fishing and environmental covariates across ecosystems; D) development of assembly rules for aggregate-species production models.  This work highlights the central role and value of datasets from multiple ecosystems, without which subsequent analyses would be impossible.