Grazing rate of zebra mussel in a shallow eutrophicated bay of the Baltic Sea

The results suggest that feeding rates of D. polymorpha are related to several environmental factors which gradients outreach the optimal range for the local mussel population. All the observed effects were non-linear with complex shapes. Variability along the resource gradient was the most important predictor of mussel feeding, followed by salinity and disturbance caused by wind. The most important interaction occurred between disturbance and resource gradient, while feeding function showed more plasticity along the latter. Mapping of environmental tipping points with the aid of machine learning methods may enable to concentrate the most relevant information about ecological functions worldwide.