Reconciling complex system models and fisheries advice: Practical examples and leads

Abstract

The move toward an ecosystem-based fisheries management (EBFM) requires new operational tools in order to support management decisions. Among them, ecosystem-and fisheries-based models are critical to quantitatively predict the consequences of future scenarios by integrating available knowledge about the ecosystem across different scales. Despite increasing development of these complex system models in the last decades, their operational use is still currently limited in Europe. Many guidelines are already available to help the development of complex system models for advice yet they are often ignored. We identified three main impediments to the use of complex system models for decision support: (1) their very complexity which is a source of uncertainty; (2) their lack of credibility, (3) and the challenge of communicating/transferring complex results to decision makers not accustomed to deal with multivariate uncertain results. In this paper, we illustrate these somehow theoretical "best practices" with tangible successful examples, which can help the transfer of complex system models from academic science to operational advice. We first focus on handling uncertainty by optimizing model complexity with regards to management objectives and technical issues. We then list up methods, such as transparent documentation and performance evaluation, to increase confidence in complex system models. Finally, we review how and where complex system models could fit within existing institutional and legal settings of the current European fisheries decision framework. We highlight where changes are required to allow for the operational use of complex system models. All methods and approaches proposed are illustrated with successful examples from fisheries science or other disciplines. This paper demonstrates that all relevant ingredients are readily available to make complex system models operational for advice.

Integrated ecological–economic fisheries models—Evaluation, review and challenges for implementation

“…The review by Lehuta et al, (2016) concentrates on methodology and model development on a subset of complex models that focus on European fisheries advice. Other types of models based on network theory such as Mullon et al, (2009) and Mullon (2013) with a global fish meal model have emerged.…”

Modeling the coastal ecosystem complex: present situation and challenges

“…The importance of ecosystem-based management has been widely recognized, in which ecosystem models are expected to play major roles (Grüss et al 2017;Kaplan and Marshall 2016;Lehuta et al 2016). Although we have discussed model development that includes detailed ecological processes suggested from observations, one problem is that this will also increase parameter uncertainty (Collie et al 2016).…”

The need for a protean fisheries science to address the degradation of exploited aquatic ecosystems

“…Such integration of all ecosystem components and their interactions offers new opportunities to explore climate change effects or those of other anthropogenic stressors on ecosystems. These models have proven useful for informing management regarding major strategic fisheries management questions (Lehuta et al 2016), such as when to fish, where to fish or which trophic levels to harvest. During the symposium several presentations highlighted the ability of trophic models to provide diagnostics of fishing impacts at the ecosystem scale for Mauritanian waters (Meissa et al, this symposium), Tunisian waters (Abdou et al 2016 However, most available tools focus on the ecological part of socio-ecosystems, and there is a lack of tools for addressing the socio-economic context in which fisheries operate (Gascuel et al 2012).…”

“…Various criteria have been proposed to seek the best model, i.e. the best trade-off between detail, simplicity, data availability and ease of communication (see for example the discussion in Cotter et al (2004) about complexity and stock assessment models and Lehuta et al 2016).…”