Evaluating biomass of Baltic filamentous cyanobacteria by image analysis

The biovolume of heterocystous cyanobacteria in the Central Baltic Sea was determined in natural, mixed-species samples by image analysis. Epifluoresc ence microscopy combined with computer-mediated image analysis was used to distinguish phycobiliprotein autofluorescent filaments of Anabaena, Aphaniz omenon and Nodularia from other phytoplankters collected on membrane filter s. A detailed description is given of the method based on the automatic mea surement of single filament morphometry. Manual interactions and mathematic al corrections were necessary during image processing to overcome filament overlapping and curvature on filters as well as the coiling of Nodularia sp p.. The autofluorescence patchiness of Aphanizomenon sp. made it necessary to trace its dimensions manually. The data obtained were used to estimate v ariations in the vertical distribution of the biomass of the 3 cyanobacteri al genera over a diel cycle.