Modelling freshwater phytoplankton communities: an exercise in validation

A total of ten validation statistics were used to test a new freshwater phytoplankton model, PROTECH-C. The biological output of the model was compared with data collected from Blelham Tarn, in the English Lake District. The usefulness of the statistics and visual fits were evaluated, as well as PROTECH-C. It was concluded that only some of the statistics were suitable, given the magnitude and range of the data, and that visual fits are still vital, particularly with ecological data which can be episodic in nature. PROTECH-C was found to simulate total chlorophyll and the algal functional groups to a satisfactory level; its failure to model individual species was expected given the ecological level at which PROTECH-C simulates, i.e. species’ populations.