Causal models as a framework for next-generation ecosystem and climate-linked stock assessments

Title: Causal models as a scientific framework for next-generation ecosystem and climate-linked stock assessments



Abstract: What do causal inference, dynamic structural equation modeling, next-generation stock assessment, and ecosystem and socioeconomic profiles have in common? They all offer tools to improve our understanding of stock productivity in a changing ocean. Here, we present the Structural Causal Enhanced Assessment Model (SCEAM), a framework that integrates causal time-series models and state-space stock assessments to quantify and forecast the effects of ecosystem and climate drivers on stock productivity. We show how causal models can reduce unexplained recruitment variance in Gulf of Alaska pollock, and improve near-term forecasts. We discuss potential onramps for operational use, including application in risk tables, hypothesis evaluation, and synthesis of process knowledge. Finally, we reflect on what happens when ecosystem linkages break down, and why building trust with stakeholders requires embracing both the promise and limits of causal reasoning.