Oral Presentation Australian Society for Microbiology Annual Scientific Meeting 2013

Molecular and morphological identification of cyanobacteria: How well do they match up? (#88)

Elvina Lee 1 , Una M Ryan 1 , Glenn McGregor 2 , Paul Monis 3 , Andrea Paparini 1
  1. Murdoch University, Murdoch, WA, Australia
  2. Department of Science, IT, Innovation and the Arts, Dutton Park, Queensland, Australia
  3. Australian Water Quality Control Centre, South Australia Water Corporation, Adelaide, SA, Australia

Cyanobacteria are ubiquitous, pleomorphic, photosynthetic microalgae. Traditionally, they have been identified and classified through morphological methods, however this process is highly subjective as distinctive phenotypic characteristics may vary within species and growth phase, or be lost as a result of environmental or culture conditions. Pleomorphism during long-term cultivation has also resulted in strains being misidentified. Hence, studies into cyanobacteria identification and diversity have increasingly incorporated both molecular and morphological methods into the process.

To determine how well these methods correlate, 37 cyanobacteria isolates were characterized both morphologically, by one or more experienced taxonomists, and by DNA sequence analysis at the 16S, rpoC1 and phycocyanin loci. Of the 37 isolates, 17 were examined by 2 independent taxonomists and results revealed that 23.5% were in agreement to the species level, 47.1% to the genus level, and the remaining 17.6% to the family or order level.  Amplification was successful for 27 isolates for 2 or more loci. Phylogenetic analysis showed that between loci, there was 25.9% agreement at the species level, 63.0% to the genus level, and 22.2% to the order level. Between morphology and phylogeny, there was only 21.6% agreement to the species level and 45.9% agreement at the genus level.  These data highlight the difficulties in identification of cyanobacteria species and the importance of using multiple loci for molecular characterization. Another important issue is the paucity of fully characterized sequences available in GenBank, which impairs the ability to accurately identify new cyanobacteria isolates.