No abstract

Background and aims -Formal inferences of the diatom phylogeny have largely depended on the nuclear-encoded small subunit of the rDNA gene (SSU). large parts of the tree remain unresolved, suggesting that new sources of data need to be applied to this question. The next largest dataset consists of the large subunit of the ribulose-bisphosphate carboxylase gene (rbcl). The photosystem II gene psbC has also been applied to problems at higher levels of the diatom phylogeny. Thus, we sequenced each of these three genes for 136 diatoms in an attempt to determine their applicability to inferring the diatom phylogeny. Methods -We attempted to obtain a more or less even sampling across the diatom tree. In particular, we increased sampling among the radial and polar centrics and among taxa that morphologically appear to be transitional between polar centrics and araphid pennates. Normal sequencing methods were used. Data were analyzed under maximum likelihood. Key results -Analysis of SSU and chloroplast data returned many of the same clades and the same general structure of the tree. Combined, the data weakly reject monophyly of the radial centrics. The chloroplast data weakly support monophyly of the polar centrics but SSU and combined data weakly reject polar centric monophyly. There may be an hitherto unrecognized clade of araphid pennates sister to the remaining pennates. Conclusion -While it is obvious that more genetic data need to be collected, perhaps the greatest obstacle to inferring an accurate, or at least global and robust, diatom phylogeny is the fact that the parts of the diatom tree that appear to be the most intractable to date (relationships among centric groups and between centrics and pennates) are also the most undersampled. This is in part due to major extinctions in the radial and polar centrics. We believe diatomists need to support more effort in both the molecular and morphological studies of these diatoms, and in the search for more information about the first half of the diatom stratigraphic record.

No abstract

No abstract

Diatoms are mostly photosynthetic eukaryotes within the heterokont lineage. Variable plastid genome sizes and extensive genome rearrangements have been observed across the diatom phylogeny, but little is known about plastid genome evolution within order- or family-level clades. The Thalassiosirales is one of the more comprehensively studied orders in terms of both genetics and morphology. Seven complete diatom plastid genomes are reported here including four Thalassiosirales: Thalassiosira weissflogii, Roundia cardiophora, Cyclotella sp. WC03_2, Cyclotella sp. L04_2, and three additional non-Thalassiosirales species Chaetoceros simplex, Cerataulina daemon, and Rhizosolenia imbricata. The sizes of the seven genomes vary from 116,459 to 129,498 bp, and their genomes are compact and lack introns. The larger size of the plastid genomes of Thalassiosirales compared to other diatoms is due primarily to expansion of the inverted repeat. Gene content within Thalassiosirales is more conserved compared to other diatom lineages. Gene order within Thalassiosirales is highly conserved except for the extensive genome rearrangement in Thalassiosira oceanica. Cyclotella nana, Thalassiosira weissflogii and Roundia cardiophora share an identical gene order, which is inferred to be the ancestral order for the Thalassiosirales, differing from that of the other two Cyclotella species by a single inversion. The genes ilvB and ilvH are missing in all six diatom plastid genomes except for Cerataulina daemon, suggesting an independent gain of these genes in this species. The acpP1 gene is missing in all Thalassiosirales, suggesting that its loss may be a synapomorphy for the order and this gene may have been functionally transferred to the nucleus. Three genes involved in photosynthesis, psaE, psaI, psaM, are missing in Rhizosolenia imbricata, which represents the first documented instance of the loss of photosynthetic genes in diatom plastid genomes.