Tryblionella persuadens comb. nov. (Bacillariaceae, Diatomeae): new observations on frustule morphology of a seldom recorded diatom

The species originally described from brackish waters of the Venetian Lagoon as Nitzschia persuadens is a diatom rarely cited in the literature since its proposition and it is here recorded for the first time in a freshwater environment in South America. Morphological features of this species, such valve slightly panduriform, with a longitudinal straight fold of the valve face, poroidal areolae, and strongly eccentric raphe system clearly assign this species to Tryblionella , and the transfer was made. Here we present new observations on the frustule morphology and comparisons with related species. Light and scanning electron microscopy data of Tryblionella persuadens comb. nov. from Cachoeira River, Northeastern Brazil are documented.


INTRODUCTION
Tryblionella W. Smith is a widespread epipelic genus, occurring in marine, brackish or high conductivity freshwaters (Round et al. 1990). This taxon was erected to generic status by Round et al. (1990), grouping the species included in four old sections from Nitzschia Hassal: Tryblionellae (W Smith) Grunow, Circumsutae Grunow, Apiculatae Grunow and Pseudotryblionella Grunow. This proposition have not been adopted by all diatomists, who considered unclear the selected discriminating criteria established (Witkowski et al. 2004). However, molecular studies performed on Bacillariaceae members have showed that Nitzschia sensu lato is not a monophyletic group, and probably should be split into several genera (Lundholm et al. 2002, Rimet et al. 2011. Rimet et al. (2011) advocated the taxonomic separation of Tryblionella and Psammodictyon D. G. Mann in Round, Crawford et Mann. (specially the later) from Nitzschia sensu stricto. Nevertheless, there are no sufficient representatives of Tryblionella included in those phylogenetic analyses for a reliable understanding about the relationships among nitzschioid genera. Whereas classification must be consistent with the phylogenetic relationships among groups, molecular tools combined with morphological features based on frustule, life form and plastids position (see Mann 1978) should clarify the genera consistency and their affiliations.
Little is known about coastal watershed diatom floras, therefore, the same occurs about possible interactions between continental and marine/ estuarine communities.
A recent floristic survey of samples collected from a coastal river in Northeast Brazil revealed a species identified as Nitzschia persuadens Cholnoky. This taxon is rarely found in the literature, and its transfer is necessary based on its affinities with other Tryblionella taxa. This species is described here, based on light and electron microscopy, representing new observations on the frustule structure and the first record to South America.

MATERIALS AND METHODS
The Cachoeira River is situated in the Eastern Basin, state of Bahia, Northeast Brazil. This coastal river is around 500 km long and has 4,600 km 2 of drainage area. Inserted into the Atlantic rainforest, it rises to 800 m above sea level, covers major urban centers and flows onto the continental shelf off Ilhéus municipality (Torres et al. 2001).
Diatoms were observed, measured and photographed with a Olympus BX-40 light microscope equipped with phase contrast and a Olympus DP-71 digital imaging system.

DISCUSSION
The species described here is clearly assigned to Tryblionella. According to Round et al. (1990), it is a genus difficult to circumscribe. The distinction among Tryblionella, Psammodictyon and Nitzschia sensu stricto is given by a combination of diagnostic features, based on valve, raphe and fibula structures. However, very eccentric raphe system accompanied by a longitudinal straight fold of the valve face is useful to Tryblionella affiliation (Mann 1978). In the species protologue, Cholnoky (1961) have already noted that T. persuadens should be included in this group (former Section Tryblionellae).  Table I. Based on type material analysis made by Krammer and Lange-Bertalot (1988) and by type illustration designated by Simonsen (1987), T. aerophila shows wider valves, broader fold, and coarser fibulae. Note that Figure 15 (Plate 51) from Krammer and Lange-Bertalot (1988), named 'N. aff. aerophila', is similar, if not identical to T. persuadens, and should not be confused with T. aerophila.
Nitzschia ligowskii, although a recently described species (Witkowski et al. 2004), used to be known by N. subconstricta Grunow, a name never validly published. It differs from T. persuadens by being longer and wider with lower striae density, and the marginal poroids are larger (Witkowski et al. 2004). Another similar species, N. buschbeckii is longer, wider, possesses a broad sternum, lower striae and fibulae densities and the poroids in the valve margins are coarser (Witkowski et al. 2004). According to our sense on phylogenetic relations among the Bacillariaceae genera, N. ligowskii and N buschbeckii should be transferred to Tryblionella.
To a lesser extent, the species mentioned above also differ by habitat. T. aerophila seems to be a typically freshwater species. Krammer and Lange-Bertalot (1988) noted that individuals of T. aerophila from brackish water bodies recorded in Central and South America and South Africa cannot be safely considered as conspecific; T. bathurstensis is from brackish, while T. sibula, N. ligowskii and N. buschbeckii are from marine habitats. T. persuadens was found in brackish (original description) and freshwater (this study) environments. Tryblionella persuadens is a poorly recorded taxon. Since its proposition, based on brackish population from Venetian Lagoon (Cholnoky 1961), the only records we have found were in Witkowski et al. (2000, p. 808 and 820), which are doubtful. Figure 7 (p. 808) showed a specimen with the following characteristics: 23.3 µm long, 10.6 µm wide, 11 fibulae and 24 striae in 10 µm, discernible poroids, which are not compatible with the original description of T. persuadens. The same occurs with figures 16 (p. 808) and 7 (p. 820), named N. cf. persuadens. Those individuals are wider (7.3, 8 µm) and show lower striae (21, 26 in 10 µm) and fibulae (10, 8 in 10 µm) densities. This is therefore the first record of this diatom in continental environments.

ACKNOWLEDGMENTS
To Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) for the scientific productivity grants to TAVL; and to Electron Microscopical Center of the Universidade Federal do Paraná for technical assistance. We are grateful to Dr. Lezilda Torgan (FZB/RS) for the availability of some important papers.