EcoState: Extending Ecopath with Ecosim to estimate biological parameters and process errors using RTMB and time-series data

• Main Authors: Thorson, James, Kristensen, Kasper, Aydin, Kerim, Gaichas, Sarah, Kimmel, David, McHuron, Elizabeth, Nielsen, Jens, Townsend, Howard, Whitehouse, Andy
• Format: Other/Unknown Material
• Language: unknown
• Published: California Digital Library (CDL) 2024
• Subjects: alaska pollock; Bering Sea; Alaska
• Online Access: http://dx.doi.org/10.32942/x2qk81

Mass-balance ecosystem models including Ecopath with Ecosim (EwE) are widely used tools for analyzing aquatic ecosystems to support strategic ecosystem-based management. These models are typically developed by first tuning unknown parameters to achieve mass balance (termed “Ecopath”), then projecting dynamics over time (“Ecosim”) while sometimes tuning predator-prey vulnerability parameters to optimize fit to available time-series. By contrast, population-dynamics (stock assessment) and multi-species models typically estimate a wide range of biological rates and parameters via their fit to time-series data, assess uncertainty via a statistical likelihood, and increasingly include process errors as “state-space models” to account for nonstationary dynamics and unmodeled ecosystem variables. Here, we introduce a state-space model “EcoState” (and associated R-package) that estimates parameters representing mass-balance dynamics directly via their fit to time-series data (absolute or relative abundance indices and fisheries catches) while also estimating the magnitude of process errors using RTMB. A case-study demonstration focused on Alaska pollock (Gadus chalcogrammus) in the eastern Bering Sea suggests that fluctuations in krill consumption are associated with cycles of increased and decreased pollock production. A self-test simulation experiment confirms that estimating process errors can improve estimates of productivity (growth and mortality) rates. Overall, we show that state-space mass-balance models can be fitted to time-series data (similar to surplus production stock assessment models), and can attribute time-varying productivity to both bottom-up and top-down drivers including the contribution of individual predator and prey interactions.