# Is there a “sweet spot” of model complexity for qualitative models used in Ecosystem-Based Management?

## TL;DR

The paper explores how complex qualitative ecosystem models should be to best support ecosystem-based management decisions.

## Contribution

The study identifies factors influencing the performance of qualitative models compared to quantitative ones in ecosystem management.

## Key findings

- Higher complexity qualitative models perform better when perturbing lower trophic level groups.
- Lower complexity models are more suitable for mid-trophic group perturbations.
- Using multiple models helps avoid misleading conclusions about ecosystem impacts.

## Abstract

Ecosystem models have been developed to help support Ecosystem-Based Management and to help provide better management advice that can account for ecosystem impacts (e.g., climate, species interactions, fishing behaviour). Quantitative end-to-end models have proven to be very useful strategically for exploring future scenarios, but are data intensive, time consuming, and require considerable expertise and training. Conversely, qualitative models have different benefits: they are less dependent on data, relatively faster to develop, can incorporate different types of information that are difficult to measure or combine, and can be co-developed with a variety of audiences. There has been an increase in the use of qualitative models for marine management, however questions have arisen about how well qualitative models perform in comparison to quantitative models, and how they can be used to inform management. Here we compare results from quantitative and qualitative ecosystem models for the same region at differing levels of model complexity to explore their relative utility for EBM. We conclude that the number of linkages between model elements and trophic position of the perturbed model were influential factors in the qualitative model behaviour. When perturbing lower trophic level groups, higher complexity models performed closer to the quantitative model. Lower complexity models were recommended when estimating scenarios with perturbations to mid-trophic groups. Careful consideration among these issues is required to develop the “sweet spot” of model complexity for qualitative ecosystem models to reflect similar results to quantitative models. In addition, utilizing multiple models to determine the strongest impacts from perturbations is recommended to avoid spurious conclusions.