The invasive cyanobacterium Cylindrospermopsis raciborskii is increasingly spreading to temperate freshwater habitats world wide and is of major concern due to its ability to produce potent toxins. It is therefore important to understand the mechanisms behind the dispersal of this species. Different hypotheses have been proposed to explain the phylogeography and mechanisms underlying the recent expansion of C. raciborskii into temperate latitudes, but there is still no conclusive evidence whether the obvious ecological success of C. raciborskii is due to selection mechanisms, physiological tolerance, climatic change or radiation after the last ice age. In this study, new isolates of C. raciborskii from Europe and Africa were genetically characterized by sequencing the ITS1, PC-IGS, nifH and rpoC1 genes and compared to corresponding sequences of C. raciborskii available in GenBank in order to test different phylogeographical hypotheses. The strains were also morphologically examined and screened for production of the hepatotoxic cylindrospermopsin (CYN). We clearly demonstrate that there are phylogenetic, morphological and toxicological differences between the isolated strains. The phylogenetic analyses revealed a clustering of the strains due to geographic origin. The ITS1 and nifH genes separated into American, European and Australian-African groups, whereas the PC-IGS and rpoC1

The aim of this study was to clarify the phylogenetic position of the three heterocystous cyanobacteria species Anabaena bergii, Aphanizomenon ovalisporum and Aphanizomenon aphanizomenoides within the order Nostocales. We determined and phylogentically analysed 16S rRNA gene and cpcBA-IGS sequences of four A. bergii, three A. ovalisporum, one A. aphanizomenoides and seven Aphanizomenon sp. strains isolated from Spain, Germany, Israel and Senegal and complemented the analyses with 2 morphometric descriptions of these strains. The phylogenetic clustering did not follow the current botanical classification. All three species clustered separately from the majority of Anabaena and Aphanizomenon strains. A. bergii and A. ovalisporum clustered close to Nodularia, whereas the position of the cluster containing the A. aphanizomenoides strain varied between the trees and the different tree constructing methods used. In addition to A. aphanizomenoides, this cluster contained the two Anabaena species A. kisseleviana and A. oumina. All three species had highly similar DNA sequences at the two fragments analysed and thus, based on evolutionary distances, might be assigned to a single species. Further, our results contradict the previously formulated suggestion that A. bergii and A. ovalisporum are 3 morphotypes of a single species. Instead, A. bergii and A. ovalisporum consistently formed separate clusters, which were less than 96.6 % similar to each other based on 16S rRNA gene sequence analysis. Our results support the idea that the taxonomy of heterocystous cyanobacteria should be revised, but also emphasize the importance of detailed morphological information when molecular data of new strains is used for taxonomy.