Development of a database system useful for identification of Anabaena spp. (Cyanobacteria)

Abstract

SUMMARYA database system useful for the identification of plank‐tonic Anabaena species based on morphology was designed. The system consisted of two database files (sample database and reference database) and three subsystems (data editing, similarity calculation, and identification). In the reference database file, 19 plank‐tonic Anabaena species were registered with their morphological characteristics of 26 attributes. The sample database of the test strain was compared with the reference one by calculating the average (St) of the similarity values estimated by peculiar methods according to the attributes. The test strain was recognized to be the same as the reference species with the highest St value. Strains of Anabaena spp. which had previously been identified were examined to test the effectiveness of the identification system. Using a non‐weighted identification system, where each morphological attribute has equal importance for identification, St (total similarity) values for the reference species previously identified were more than 0.76; however, they were not always the highest value obtained. Using a weighted identification system, where trichome form, trichome shape and akinete morphology were given greater importance in the identification subsystem, St values for the reference species to which the test strains had previously been assigned were highest. This system proved useful for identification of Anabaena species, and has the additional advantage that there are no individual differences in the diagnosis when different users interrogate the system.