Predictive modelling of seabed habitats: case study of subtidal kelp forests on the coast of Brittany, France

Predictive modelling to map subtidal communities is an alternative to “traditional” methods, such as direct sampling, remote sensing and acoustic survey, which are neither time- nor cost-effective for vast expanses. The principle of this modelling is the use of a combination of environmental key parameters to produce rules to understand species distribution and hence generate predictive maps. This study focuses on subtidal kelp forests (KF) on the coast of Brittany, France. The most significant key parameters to predict KF frequency are (1) the nature of the substrate, (2) depth, (3) water transparency, (4) water surface temperature and (5) hydrodynamics associated with the flexibility of algae in a flow. All these parameters are integrated in a spatial model, built using a Geographical Information System. This model results in a KF frequency map, where sites with optimum key parameters show a deeper limit of disappearance. After validation, the model is used in the context of Climate Change to estimate the effect of environmental variation on this depth limit of KF. Thus, the effects of both an increase in water temperature and a decrease in its transparency could lead to the complete disappearance of KF.