MOLECULAR PHYLOGENY OF SELECTED MEMBERS OF THE ORDER THALASSIOSIRALES (BACILLARIOPHYTA) AND EVOLUTION OF THE FULTOPORTULA¹

Abstract

Recent phylogenetic studies of the diatoms indicate that members of the order Thalassiosirales occupy an interesting position in the diatom evolutionary tree. Despite their radial morphology and scaly auxospores, they are consistently recovered in molecular analyses as a member of subdivision Bacillariophytina and a sister clade to non-fultoportulate and non-radial lithodesmioids. This study included 46 species from nine traditionally accepted extant genera, and analyzed 43 nuclear small subunit (SSU) rRNA sequences in parallel with a survey of the variation in fultoportula structure. Three possible scenarios leading to the evolution of the fultoportula are discussed in the context of molecular and morphological similarities between the examined Thalassiosirales and their SSU rRNA sister clade Lithodesmiales. We speculate that the fultoportula might be derived by a modification of either a cribrum in an areola (fultoportula within an areola), or structures similar to marginal ridges now seen in lithodesmioids around a cluster of poroids (fultoportula in a tube), or finally, that the central fultoportula may have an origin different from the marginal fultoportulae. Our data confirm that fultoportula-bearing diatoms constitute a natural phylogenetic group. The families Thalassiosiraceae, Skeletonemaceae, and Stephanodiscaceae and the genus Thalassiosira Cleve were unexpectedly found to be paraphyletic. Further, Cyclotella Kutz. and Stephanodiscus Ehr. may not be closely related and some species of these genera are more closely allied to other species of Thalassiosira. The generitype, T. nordenskioeldii, is embedded within a large poorly structured cluster of species that includes several members of Thalassiosira, Planktoniella sol, Minidiscus trioculatus, and two members of Stephanodiscus. An emendment of the order Lithodesmiales and the family Lauderiaceae are proposed.

Evolution of the diatoms: insights from fossil, biological and molecular data

“…the Lithodesmiales and the Thalassiosirales. However, such discoveries prompt new and more detailed structural and cytological investigations to reveal whether the differences are more apparent than real and to understand their basis (Kaczmarska et al 2006). We still are lacking much detail about the life cycle and protoplast structure in many diatom genera, so that some of the sister relationships recovered in the molecular analyses remain 'bald' cladistic hypotheses, with little biological meaning.…”

“…This fact led Medlin et al (1993) and Medlin & Kaczmarska (2004) to suggest that the central strutted process could be homologous with the bilabiate process of the Lithodesmiales. Kaczmarska et al (2006) have suggested further that the external tubes of the marginal strutted processes could be formed by a fragmentation, then encircling and fusing of the marginal fringe in the Lithodesmiales to form hollow tubes above the pore structure of the fultoportula. Two different types of strutted processes are known in the Thalassiosirales (Kaczmarska et al 2006), but they could all be derived in this fashion.…”

“…Kaczmarska et al (2006) have suggested further that the external tubes of the marginal strutted processes could be formed by a fragmentation, then encircling and fusing of the marginal fringe in the Lithodesmiales to form hollow tubes above the pore structure of the fultoportula. Two different types of strutted processes are known in the Thalassiosirales (Kaczmarska et al 2006), but they could all be derived in this fashion. Thus, from this scenario, perhaps the central strutted process is evolutionarily distinct from the marginal ones.…”

“…In a molecular analysis of all extant genera of Thalassiosirales (Kaczmarska et al 2006;Fig. 74), Porosira E. Jørgensen and Lauderia Cleve root the lineage.…”

Ultrastructure of a Hyalodiscus species (Bacillariophyceae; Subclass: Coscinodiscophycidae, Fam. Hyalodiscaceae) from brackish waters of Tonga, Oceania.

“…The genus Hyalodiscus contrast clearly to the more derived genus Coscinodiscus which has two types of rimoportulae instead of one and a more complex structure of loculi (large and small ones). Both more primitive families Hyalodiscaceae and Coscinodiscaceae are differentiated from the more advanced Thalassioraceae by missing of fultoportulae (KaczMaRSKa et al 2006). In spite of the generally different frustule geometry of Hyalodiscus (short pervalvar axis) and Podosira (extended pervalvar axis) ultrastructural similarities are given especially in the type of rimoportulae (labiate processes), but their densities differ considerably (much less dense in Hyalodiscus than Podosira see Round et al 1990).…”