Original paper

Revisionary concepts of species in the Cyanobacteria and their applications

Castenholz, Richard W.; Norris, Tracy B.


This paper discusses the various concepts of species and genus and also suggests future usage of these two terms. Cyanobacterial taxonomy must hereafter utilize a polyphasic approach that includes at least some elements of genetics, morphology, cytology, ultrastructure, physiology, biochemistry and ecology - otherwise, descriptions and names, both past and future become nearly meaningless. Using phenotypic characters alone to describe and name taxa is no longer a legitimate pursuit, since it is well known that cyanobacteria of similar morphology and physiology may be quite different genetically, and cyanobacteria of the same genetic makeup may appear quite different under different ecological or physiological conditions, the latter situation often resulting simply from differential gene expression. In genetic terms, species with precise boundaries probably do not exist, but are merely peaks that stand out in never-ending variations in sequences of single or multiple genes or of complete genomes. The central question is whether it will be possible to resolve these species peaks and what is the appropriate level of genetic information to do so (several genes or complete genomes?). One of the main problems with even describing new species or genera in both the traditional or bacteriological sense is that the delineation of new cyanobacteria using genetic and phenotypic criteria is escalating at such an overwhelming rate that the number of cyanobacterial taxonomists who would actually write a formal species (or genus) description would be inadequate to cope with the burgeoning number of strains being characterized and entered into databases. There will be numerous and nevernding revisions, and a new species or genus being described will eventually find itself closely allied to a number of new variants unknown at the time of the formal description. Until there are complete genomes published for hundreds or thousands of cyanobacterial isolates, it may be impossible to determine the boundaries of species or genera, and even then the boundaries may be arbitrary and decided simply by consensus. For now we believe that the current rules of nomenclature (Botanical and Bacteriological) should be used when formal descriptions of very unique taxa are prepared, but that another form of delineating new taxa should now be used for the burgeoning number of strains being discovered. Temporarily, at least, "traditional" and more recent generic names should be used with strain numbers for each clonal isolate or collected specimen, dispensing with species epithets (as in the Pasteur Culture Collection). In some cases the name of the closest morphological equivalent should be given (using cf. from Latin confer = compare) in order to aid future taxonomists. Cyanobacteriologists need to continue to monitor the rapidly increasing genetic and phenotypic information and wait until some logical and pragmatic pattern emerges before attempting to specifically delineate the boundaries of natural species and genera and perhaps, even then, creating new operational taxonomic units that replace the traditional binomials. Polyphasic taxonomy is advancing rapidly, and several examples are presented. The availability of cyanobacteriaspecific primers for genetic identification and the fact that morphological, ecological and physiological information may be obtained without pure cultures, means that new taxa can be both described and labeled quite adequately in a manner already widely in use (e.g. Synechocystis PCC 6803).


cyanobacteriaspecies concepttaxonomypolyphasic methodecotypegenotypephenotype