‘Building the (Im)perfect Beast’: Strategies for Identifying Appropriate Spatial Stock Assessment Model Complexity From an International, Blinded High‐Resolution Simulation Experiment

Details

• Journal Title:
  Fish and Fisheries
• Personal Author:
  Berger, Aaron M. ; Goethel, Daniel R. ; Hoyle, Simon D. ; Lynch, Patrick ; Barceló, Caren ; Dunn, Alistar ; Langseth, Brian J. ; Minte‐Vera, Carolina ; Day, Jemery ; Xu, Haikun ; Izquierdo, Francisco ; Fu, Dan ; Ducharme‐Barth, Nicholas D. ; Vincent, Mathew ; Grüss, Arnaud ; Deroba, Jonathan J. ; Correa, Giancarlo M. ; McKenzie, Jeremy ; Butler, Will ; Cao, Jie ; Marsh, Craig ; A'mar, Teresa ; Bartolino, Valerio ; Cardinale, Massimiliano ; Castillo‐Jordan, Claudio ; Elvarsson, Bjarki Þór ; Hampton, John ; Havron, Andrea ; Mace, Pamela ; Magnusson, Arni ; Maunder, Mark ; Methot, Richard ; Mormede, Sophie ; Pennino, Maria Grazia ; Perez‐Rodriguez, Alfonso ; Cousido‐Rocha, Marta ; Teears, Thomas ; Urtizberea, Agurtzane
• NOAA Program & Office:
  NMFS (National Marine Fisheries Service) ; AFSC (Alaska Fisheries Science Center) ; OST (Office of Science and Technology) ; NEFSC (Northeast Fisheries Science Center) ; NWFSC (Northwest Fisheries Science Center) ; PIFSC (Pacific Islands Fisheries Science Center) ; SEFSC (Southeast Fisheries Science Center)
• Description:
  Despite their potential to inform sustainable regional harvest and climate-resilient fisheries management, spatial stock assessment models remain underused for management advice. To identify barriers that inhibit broader use of these methods, we conducted a blinded international simulation experiment mimicking real-world stock assessment development when confronting spatial complexity. Seven analyst teams built spatially aggregated and spatially explicit assessment models using data simulated from high-resolution operating models based on Indian Ocean yellowfin tuna and Ross Sea Antarctic toothfish dynamics. Each team documented how assessment software platform, data analyses, model building approach, and diagnostics influenced model complexity and realism. A consensus emerged on key assessment building approaches: (1) conduct high-resolution data analyses to identify appropriate spatial structure; (2) start with simplified models and incrementally add complexity; (3) iteratively evaluate diagnostics to determine necessary spatial complexity; and (4) maintain models with different spatial structures to aid interpretation. The experiment also revealed several valuable insights for parameterising assessments, including consideration of data pre-processing with spatiotemporal models to better inform data-sparse regions; regression trees to identify fleet and spatial structure; trade-offs in complexity between productivity and movement dynamics to achieve tractable and stable model structures; and ensemble modelling approaches to address structural uncertainty. Our findings demonstrate that international collaborations and simulation experiments are crucial for addressing challenges in implementing spatial stock assessments and for evaluating whether their added complexity is justified given management objectives. Broader collaborations are encouraged to foster innovation in fisheries management and to help recognise the practical trade-offs between model parsimony and complexity.
• Source:
  Fish and Fisheries (2025)
• DOI:
  https://doi.org/10.1111/faf.70048
• ISSN:
  1467-2960 ; 1467-2979
• Format:
  pdf
• Publisher:
  Wiley
• Document Type:
  Journal Article
• License:
  https://creativecommons.org/licenses/by-nc/4.0/
• Rights Information:
  CC BY-NC
• Compliance:
  Submitted
• Main Document Checksum:
  urn:sha-512:a302b02b1e3becc421b7e2fd863142edf68dbf6efbbfbceeebda6ed610ed57a20ae1edd583d3c8386fec3c81498b6948310349f11076fdfd52cc6dac39d6cca3
• Download URL:
  https://repository.library.noaa.gov/view/noaa/72345/noaa_72345_DS1.pdf
• File Type:
  [PDF - 1.37 MB ]