Applying essential ecosystem service variables to analyse thirty years of wild salmon provisioning trends in Canada

Flavio Affinito , Marie-Josee Fortin, Andrew Gonzalez 

Systematic monitoring of ecosystem services (ES) is crucial for achieving sustainability goals but hampered by siloed and disjointed monitoring efforts that rarely consider the full range of social, economic, and ecological variables shaping ES dynamics. The essential ecosystem service variables (EESV) framework is intended to tackle these challenges but its application remains limited. Using British Columbia's wild Pacific salmon commercial fishery as a case study, we conduct the first operationalization of the EESV framework, integrating 27 years of monitoring data from diverse sources to assess change in the ES provided by wild Pacific salmon fisheries in a multi-species analysis conducted at the provincial scale of BC. We develop and test a causal model incorporating five essential variables (salmon abundance, fishing effort, salmon catch, landed value, and market demand) and seven drivers, including sea surface temperature, hatchery releases, and fishing licenses. Our results show that rising sea surface temperatures negatively impact salmon returns, while population enhancement through hatcheries has generally weak effects, with some evidence suggesting that they may be causing declines in abundance. We also evaluate the government’s license retirement program, finding limited success in reducing fishing effort and improving the industry’s financial viability. Bayesian modeling highlights greater uncertainty in ecological versus socio-economic predictions, suggesting an opportunity for better monitoring. Crucially, the inability to operationalize the relational value EESV class underscores limitations in current monitoring programs. Overall, our findings show market forces sustaining ES value despite declining effort and uncertain ecological dynamics. By demonstrating the EESV framework's utility for integrating data into a causal representation and diagnosing monitoring system limitations, this study advances the methodology needed for standardized ES monitoring in the case of provisioning services.