Towards a comprehensive framework for providing management advice from statistical inference using population dynamics models

Abstract

There has been substantial progress in fitting population dynamics models to data and this has greatly improved management advice in a variety of situations from exploitation to conservation. One of the major developments has been integrated analysis where multiple diverse data sets are fit simultaneously within the same model. However, issues such as model misspecification, unmodelled process variation, and data weighting make integrated analysis problematic. Here I provide a personal perspective on a framework for Model Development (FMD) based on the Center for the Advancement of Population Assessment Methodology (CAPAM) workshops and special issues, my own research, and other information. The FMD is motivated by fisheries stock assessment but is relevant to any form of population dynamics modelling or modelling in general. I provide an outline of the modeling framework and discuss the important topic of data weighting. The FMD starts with one or more conceptual models which are implemented as population dynamics models fit to data using a comprehensively researched Good Practices Guide (GPG). The models are evaluated, improved, and selected, based on a diagnostic “expert” system that has been rigorously developed using a comprehensive simulation analysis. The final models that are accepted in the ensemble are equally weighted (until the data weighting issue is fully resolved) to provide management advice. I also outline necessary future research.