Abstract
As climate change is having increasingly visible impacts on coastal regions, it is urgent to better understand its effects on the state of ecosystems and the services they provide. To assess the direction and magnitude of change in the high-use waters of the Sept-Îles region in the Gulf of St-Lawrence, we sampled 35 sites along the coast and explored the relationship between surface diatom assemblages and 21 environmental variables. Temperature (14.62%), Secchi depth (12.97%), dissolved silica (6.91%), and magnesium (6.23%) were determined to be significant and independent variables explaining variation in diatom distribution, together explaining 28.6% of the assemblage variance. Temperature and Secchi depth met the λ(1)/λ(2) > 0.5 criterion, but only temperature was selected to develop a transfer function, as these two variables were strongly correlated. A diatom-based temperature transfer function was then developed using weighted-averaging partial least squares (WAPLS; 3; r(2) = 0.59, RMSEP = 0.45°C), based on a training set of 32 sites and 55 benthic taxa. However, the model exhibited sensitivity to spatial autocorrelation and may have been influenced by secondary and unmeasured variables. Despite these limitations, the model provides valuable insight into the spatial variability of diatom assemblages and offers potential for paleoenvironmental reconstructions. For optimal inferences, the model should be applied within the studied area. This study contributes to understanding how coastal diatom assemblages respond to environmental gradients and highlights the importance of diatom-based monitoring to assess ongoing ecological changes in aquatic environments.