From calibration to identifiability analysis: Using a multi-reserve DEB model to understand holopelagic Sargassum spp. growth

Unprecedented quantities of the brown floating macroalgae Sargassum have been recorded in the tropical North Atlantic Ocean. However, the environmental factors affecting holopelagic Sargassum spp. growth and survival remain poorly understood. Here, we developed a multi-reserve DEB model for Sargassum spp. to link environmental conditions (temperature, nutrients, and light) to the organism physiology. In the context of DEB theory, estimating a multi-reserve model parameters can be particularly challenging and strongly relies on the availability of the datasets with information about physiological processes. We conducted a literature review of Sargassum spp. physiological responses, and although some data have been reported, experimental datasets on some key physiological processes are still relatively scarce. From the available data we performed an estimation of the DEB model parameters. We then followed a “virtual ecologist” approach to create a framework linking modelers and empiricists, to analyze which experiments that are most needed to reduce parameter uncertainty. We tested parameter identifiability by creating “virtual” observed data and we analyzed the type of datasets that reduce the uncertainty on parameter estimates, thus increasing their identifiability and the prediction power of the algae model. Our analysis showed to which extent new experiments on nutrient uptake and nutrient limitation would reduce the uncertainty of key model parameters and improve our understanding of holopelagic Sargassum spp. physiology. More generally, our study demonstrates how the use of DEB models, coupled with parameter identifiability analyses, can help design most needed experiments thereby contributing to an integrated approach between experiments and modeling. • Multi-reserve DEB model captured how environment affects holopelagic Sargassum growth. • Review showed diversity of physiological responses studied of holopelagic Sargassum. • We identified data that improves DEB parameters estimation and predictability. • Creation of approach to help design targeted experiments to estimate model. parameters.