A general framework for combining ecosystem models

Michael A Spence,Robert B Thorpe,Natalia Serpetti,Paul G Blackwell,Steven Mackinson,Johanna J Heymans,Axel G Rossberg,Michael R Heath,Julia L Blanchard,Douglas C Speirs

doi:10.1111/faf.12310

Abstract

AbstractWhen making predictions about ecosystems, we often have available a number of different ecosystem models that attempt to represent their dynamics in a detailed mechanistic way. Each of these can be used as a simulator of large‐scale experiments and make projections about the fate of ecosystems under different scenarios to support the development of appropriate management strategies. However, structural differences, systematic discrepancies and uncertainties lead to different models giving different predictions. This is further complicated by the fact that the models may not be run with the same functional groups, spatial structure or time scale. Rather than simply trying to select a “best” model, or taking some weighted average, it is important to exploit the strengths of each of the models, while learning from the differences between them. To achieve this, we construct a flexible statistical model of the relationships between a collection of mechanistic models and their biases, allowing for structural and parameter uncertainty and for different ways of representing reality. Using this statistical meta‐model, we can combine prior beliefs, model estimates and direct observations using Bayesian methods and make coherent predictions of future outcomes under different scenarios with robust measures of uncertainty. In this study, we take a diverse ensemble of existing North Sea ecosystem models and demonstrate the utility of our framework by applying it to answer the question what would have happened to demersal fish if fishing was to stop.

Highlights

Ecosystem models are widely used to support policy decisions, including fisheries and marine environmental policies (Hyder et al, 2015)
As we are interested in measuring uncertainty, our statistical modelling will apply Bayesian inference methods (Robert, 2007), and our analysis will consider any relevant prior knowledge as well as simulator outputs that predict what would happen in the future under different management scenarios
To describe the relationship between the simulators and the truth, we developed an ensemble model that describes the population of simulators, its dynamics and its relation with the true quantity of interest

Summary

| INTRODUCTION

Ecosystem models are widely used to support policy decisions, including fisheries and marine environmental policies (Hyder et al, 2015). An alternative approach is to try and find the “best” model(s) (Johnson & Omland, 2004; Payne et al, 2015) These methods imply that at least one of the models is “correct,” in the sense that it can predict the true output. It is important to take these similarities into account rather than treating the models as independent (Rougier, Goldstein, & House, 2013) Another approach is to think of the ecosystem models as coming from a population of such models (Chandler, 2013; Leith & Chandler, 2010; Tebaldi & Sansó, 2009) and describe how the population differs from reality. We describe an ensemble model which is based on the principles of Chandler (2013) but which models the outputs themselves, varying in form between the different ecosystem

INTRODUCTION

| Results

| DISCUSSION

| Future work and extensions

| Conclusion