Differentiating Iconella from Surirella (Bacillariophyceae): typifying four Ehrenberg names and a preliminary checklist of the African taxa

Abstract To comply with the new phylogeny within the Surirellales as supported by molecular and morphological data, re-evaluations and re-combinations of taxa from and within the genera Surirella, Cymatopleura, and Stenopterobia and with the re-established genus Iconella are necessary. Since the African diatom flora is rich with taxa from these genera, especially Iconella, and the authors have studied these taxa recently, describing also new taxa, a preliminary checklist of African Iconella and Surirella is here presented. 94 names are contained on this list. 57 taxa have been transferred to Iconella; 55 taxa were formerly ranked within Surirella and two taxa within Stenopterobia. 10 taxa have stayed within Surirella and six taxa have been transferred from Cymatopleura to Surirella. 20 Surirella and 1 Stenopterobia names are listed which are either unrevised or unrevisable since morphological data is missing. Four names and taxa described by Ehrenberg are here typified. Two had been transferred to Iconella already: Iconella bifrons (Ehrenb.) Ruck & Nakov and Iconella splendida (Ehrenb.) Ruck & Nakov. Two are re-transferred from Cymatopleura to Surirella: Surirella librile (Ehrenb.) Ehrenb. and Surirella undulata (Ehrenb.) Ehrenb.; both taxa are currently known by their younger synonyms: Cymatopleura solea (Bréb.) W. Smith and Cymatopleura elliptica (Bréb. ex Kützing) W. Smith. Lectotypes for Iconella bifrons, I. splendida, Surirella librile, and S. undulata were designated.


Introduction
Surirella taxa have been recognized, drawn, and described very early in diatom history since they often have large cells. The genus Surirella Bory is within the first published diatom genera which are still in current use: Bacillaria by Gmelin  The genus name Surirella was introduced by P.J.F. Turpin in 1828 who had found it in a collection by the French medical doctor Suriray from brackish waters at the coast of Le Havre in France. He published beautiful drawings which had been enlarged in the microscope by 300×. Ehrenberg also used this 300x enlargement for his research and used this genus name first in 1834 for Surirella bifrons and Surirella splendida; in his 1838 publication (Ehrenberg 1838) he ranked Surirella as a subgenus of Navicula and contained in it the species librile, splendida, bifrons, undulata, striatula (type of the name of the genus Surirella introduced by Turpin), and constricta (no Surirella according to Jahn and Kusber 2004). For each of these he added a ? between the genus and the epithet which meant that he thought that this species might belong to a new genus to be differentiated from Navicula; at the end of the text he wrote that they definitely belong to the genus Surirella because of their different mode of division in comparison to Navicula. By 1845 Ehrenberg (1845a, b) had also recombined Navicula librile and Navicula undulata with Surirella (see typifications below).
Subsequently, more Surirella taxa were discovered. W. Smith (1851: 7) explains the morphology of Surirella: "Valves concave, with a longitudinal central line and margins produced beyond the suture (winged). … The concavity of the valves, their winged margins, and the longitudinal central line, which wants the central depression so conspicuous in the Naviculae, are characters which sufficiently distinguish Surirella from all other genera. I believe a careful examination of the loricae … would detect the presence of alae in all the species." In this paper he also described and differentiated his new genus Cymatopleura against Surirella, the main differences being "the undulated surface of the valves seems to indicate a peculiarity of structure sufficient to constitute a generic difference, and the absence of alae and costae implies a further diversity in the internal character which cannot be regarded as unimportant" (W. Smith 1851: 12). Subsequently, W. Smith recombined Cymatopleura solea (= S. librile Ehrenb.) and Cymatopleura elliptica (= S. undulata Ehrenb.). In his Treatise on the Diatoms, Van Heurck (1896: 374) reintroduced and validated the genus Stenopterobia which had been first mentioned by Brébisson; his short differential diagnosis against Surirella is: "Frustules very elongated and very narrow, sometimes sigmoid." All the above mentioned genera, Surirella, Cymatopleura, Stenopterobia, Campylodiscus (for C. clypeus (Ehrenb.) Ehrenb. ex Kütz. see Poulíčková and Jahn 2007) are part of the order Surirellales (sensu Round et al. 1990, Ruck and Kociolek 2004, and Ruck et al. 2016a which are canal-raphe-bearing diatoms with a circumferential raphe at the entire valve margin. The genera Epithemia and Rhopalodia which have a canal-raphe-not positioned around the entire valve margin, had been placed into the order Rhopalodiales (Round et al. 1990) but Ruck et al. (2016a) placed them also into the order Surirellales because their monophyly is strongly supported by molecular data Theriot 2011, Ruck et al. 2016a). However, the publications of Ruck et al. (2016aRuck et al. ( , 2016b, performed with morphology and molecular markers on those Surirellales, strongly reject the monophyly of several genera in the current classification (Round et al. 1990), especially concerning the genera Surirella and Campylodiscus. In order to provide a home to taxa which do not fit into their strict genus definition, Ruck et al. (2016b) reintroduced the genus Iconella which had been established by Jurilij in 1949 and Coronia which had been established as a subgenus by Ehrenberg, validated by Grunow and raised to genus rank by Ruck and Guiry (2016).
In the tropical African aquatic ecosystems, taxa from the genera Surirella and Cymatopleura, as traditionally known, play an important role (Ross 1983, Cocquyt and Vyverman 1994, Cocquyt 2000. In typifying historical material from African waters as described by Otto Müller (Cocquyt and Jahn 2005, 2007a, 2007b, 2007c, 2007d, by Cholnoky (Cocquyt et al. 2017), by Foged (Cocquyt and Kusber 2010), by Woodhead and Tweed (Cocquyt et al 2013), we have tried to reevaluate earlier findings of these taxa as well as their endemism. In order to help researchers to name their taxa correctly, we are providing a list of African taxa which have been recombined with a different genus; we are also listing those taxa whose names did not change. Since some of Ehrenberg's species have been the basis for varieties of African taxa, we are including the typification of four taxa originally described by Ehrenberg and synonymizing two younger taxa.

Material and methods
From the Ehrenberg Collection at BHUPM (Museum für Naturkunde, Berlin), the following materials (for details of the collection see Jahn and Kusber 2004)  New names and typifications are registered in PhycoBank ), a registration system for nomenclatural acts (see Barkworth et al. 2016) which is currently in the trial phase. Stable http identifiers are linking to the prototype portal. When possible, we are using long-term stable and semantic web compatible identifiers for specimens according to .
Taxonomical comment. Ehrenberg (1832) published Navicula librile by a description which included the length of 1/10 Paris Line which is 225.6 µm. But this measurement does not correspond to the first observations he made in Berlin 1826 drawn on a small piece of paper (Fig. 3E) and glued onto the drawing sheet BHUPM 1151. Nevertheless, the published measurement corresponds perfectly to two of his specimens on his drawing sheet BHUPM 1151 showing a living cell in valvar and girdle view ( Fig. 3A-B). Therefore, Ehrenberg (1832) was the first who described the species which was some years later described again as Cymbella solea Bréb. & Godey (1835) which was later recombined as Cymatopleura solea (Bréb.) W. Sm. (1851) and became type of the name of the genus Cymatopleura (Smith 1851). Ehrenberg's specimens, probably deposited in 1835 or 1836 (see Ehrenberg 1838) give proof ( Fig. 3C-D) of his earlier findings (Ehrenberg 1832). In addition, Ehrenberg apparently also observed the form which is identified today as "Cymatopleura solea var. apiculata" (cf. Fig. 3F, e.g. Krammer & Lange-Bertalot 1988, Hofmann et al. 2013. Schoeman and Archibald (1979) had accepted Ehrenberg's taxon as having priority under Cymatopleura. Later Cymatopleura was conserved against Sphinctocystis Hassall with Cymatopleura solea as its type (see Wiersema et al. 2015). Since Cymatopleura is here not accepted at the rank of a genus, this conservation is not applicable to our taxonomic treatment.  Ehrenberg (1845b) introduced and used the name Surirella undulata. In this publication he described all species new to science formally with a Latin diagnosis. Because he did not mark the species as new to science, Ehrenberg introduced the name Surirella undulata as a new combination of Navicula undulata under the then accepted genus name Surirella. This combination can be verified by the drawing Ehrenberg (1854) provided e.g. for Berlin material "Brakisches, strichweis lebendes, Erdlager unter Berlin" (Ehrenberg 1854: pl. 14: fig. 39). Since Ehrenberg published this taxon name already in 1838, his name has priority over Surirella elliptica.

Autapomorphies
In the Surirellaceae the raphe canal runs marginally at the edge of the valve. This canal is interrupted on the external valve face at the poles of the valve while internally the raphe is continuous at the head pole, and interrupted at the base pole. Differences between the three genera had been defined as (according to Hofmann et al 2011): • Cymatopleura: valves are crossed by several large undulations which are not interrupted near the median line (= axial area). The raphe is located within a shallow keel (Spaulding and Edlund 2008). • Stenopterobia: valves are elongated or curved sigmoid-like with equally sized poles.
The canal raphe is raised above the valve onto a keel (Spaulding and Edlund 2010). • Surirella: valves are iso-or heteropolar, transapical undulations are finely structured and interrupted near the median line.


Pinnatae group: raphe canal sits directly at the valve mantle; the raphe is interrupted at both poles. Supporting elements are the fibulae which project from the valve mantle more or less into the center of the valve face.
 Robustae: raphe canal rises above valve face and mantle and is located on a wing. Where the canals of the wings, the alar canals, meet the valve face, in LM appears an apically running wavy line which has been named a loop (Schleifenbildung). Between the alar canals lie fenestrae.
These traditional differentiations based on outline, undulations and median line (formerly named pseudoraphe or axial area) were not supported by the molecular data (Ruck et al. 2016a). Ruck et al. (2016a) therefore proposed morphological autapo-morphies for the differentiation of genera. As a true autapomorphy they accepted only the morphological differentiation between the Pinnatae and the Robustae group within Surirella which means the raphe canal is located either directly on the mantle (Pinnatae) or rises above the valve and mantle and has alar canals with fenestral openings occluded by fenestral bars (Robustae).
Since the type of the name of the genus Surirella, S. striatula, belongs to the Pinnatae group, the Pinnatae make up the true Surirella genus including also the taxa from the Cymatopleura genera because their raphe canal also is located on the valve mantle. Taxa from the Robustae group as well as Stenopterobia taxa -and a few Campylodiscus taxa i.e. C. hibernicus -belong to the reinstated genus Iconella. Since alar canals have also been found in marine Campylodiscus sensu lato (now Coronia (Ehrenb. ex Kütz.) Ruck & Guiry), an additional autapomorphy for Iconella besides the occluded fenestral openings are the internally rimmed pores.
This means that the above list of features for identifying the genera needs to be revised (according to Ruck et al. 2016): • Campylodiscus s.s. (C. clypeus only plus formerly Surirella Fastuosae; most of its marine taxa are now Coronia, the freshwater taxa Iconella): communication between the raphe canal and interior through a funnel-or chalice-shaped structure. • Coronia (formerly marine Campylodiscus): raphe canal rises above the valve and mantle; it has alar canals with fenestral openings often unoccluded and with simple unrimmed pores. • Surirella s.s. (restricted to the Surirella Pinnatae plus Cymatopleura): the raphe canal is located directly on the mantle. • Iconella (formerly Surirella Robustae, Stenopterobia plus formerly Campylodiscus Robusti): raphe canal rises above the valve face and mantle and has alar canals with fenestral openings occluded by fenestral bars with internally rimmed pores.
Campylodiscus taxa reported from tropical Africa are few. Beside the more common C. clypeus and C. clypeus var. bicostata (W. Sm. ex Roper) Hust. the only endemic species is Campylodiscus tanganicae Hust., reported from Lake Tanganyika. Since we cannot determine currently to which genus the African taxa associated historically with Campylodiscus belong, we have excluded them from this study. Marine Coronia taxa are also not part of this study.
The African Rhopalodia and Epithemia taxa as described in O. Müllers papers are currently being studied by us and will be published elsewhere.
=  fig. 15 fits the criterium "illustration with analyses" (McNeill et al. 2012, Art. 38.10) because many small spinules on the valve surface are clearly shown. Therefore we have choosen pl. 23: fig. 15 as the lectotype. fig. 16 is less detailed. We exclude the depicted specimen collected at Khayenmatay ( fig. 17) from the species because with its denser costae and less distinct wing projection it probably belongs to a different species. Comment. Type specimen not studied but specimens from Lake Tanganyika observed (Cocquyt 1998).
Comment. For taxonomical results and discussion see . Conclusion 55 taxa -formerly ranked within Surirella -have been transferred to Iconella; most of these have been shown to be endemic (Ross 1983, Cocquyt and Vyverman 1994, Cocquyt 1998, 2000 and many of them, especially the large species, have become planktonic in the East African great lakes (Müller 1905, Hustedt in Huber-Pestalozzi 1942, Cocquyt 1998). In addition, two taxa -formerly ranked within Stenopterobia -have been transferred to Iconella. 10 taxa have stayed within Surirella, (although the position of S. sparsipunctata has to be genetically verified), and six taxa have been transferred from Cymatopleura to Surirella. For completeness sake, 21 taxa have been listed which are either unrevised or unrevisable because missing morphological data do not allow us to decide if the raphe is raised on a keel. When more taxa from the genera Iconella and Surirella have been studied molecularly, especially the endemic species from Africa and other tropical regions, further autapomorphies might be discovered which might support the differentiation into further groups. With the currently available data, the solution by Ruck et al (2016a, b) clarifies their phylogeny and presents a very workable approach.