Odontidium hyemale (Roth) Kützing (1844: 44

Jüttner, Ingrid, Williams, David M., Levkov, Zlatko, Falasco, Elisa, Battegazzore, Maurizio, Cantonati, Marco, Vijver, Bart Van De, Angele, Catalena & Ector, Luc, 2015, Reinvestigation of the type material for Odontidium hyemale (Roth) Kützing and related species, with description of four new species in the genus Odontidium (Fragilariaceae, Bacillariophyta), Phytotaxa 234 (1), pp. 1-36 : 9-13

Odontidium hyemale (Roth) Kützing (1844: 44 , pl. 17, fig. 4) ( Figs 8–60 )

Basionym: Conferva hyemalis Roth (1800a: 506)

Synonyms: Fragilaria hyemalis (Roth) Lyngbye (1819: 185 , tab. 63, E 1–6), Candollella hyemalis (Roth) Gaillon (1833: 34) , Temachium hiemale (Roth) Wallroth (1833: 116) , Lysigonium hyemale (Roth) Trevisan (1848: 96) , Diatoma hyemale (Roth) Heiberg (1863: 58) , Bacillaria hiemalis (Roth) Tömösváry (1879: 150) , Neodiatoma hiemale (Lyngbye) Kuntze (1891: 906) .

LM ( Figs 8–52 ): Frustules rectangular in girdle view ( Figs 32–35 ), valves linear-elliptic to elliptic-rhombic, small valves elliptic-rhombic or rhombic, with rounded apices, length 16.1–58.1 μm, width 8.3–11.4 μm, length/width ratio 1.9–6.2 ( Figs 8–32 , 36–52 ). Valve margins of larger cells more or less parallel, tapering towards apices in the distal quarter of each valve; valve margins in smaller valves tapering from valve centre. Transapical ribs nearly all primary, perpendicular or at slight angle to sternum; 3–5 in 10 μm. Striae between ribs appear parallel, 1–10 rows between pairs of ribs. Cingulum composed of many bands. SEM ( Figs 53–60 ): External valve face slightly rippled with or without small spines (spinules), located along valve margin including pore fields, upper part of mantle and near margin on valve face ( Figs 53, 54, 56, 57, 59 ). Spines irregularly positioned on virgae protruding in variable directions ( Fig. 59 ). Virgae of varying widths, vimines much shorter ( Figs 54, 56 ). Striae uniseriate, parallel to slightly radiate towards poles. Sternum relatively narrow and slightly irregular, in some valves extending into apical pore field ( Figs 54, 56 ). One rimoportula per valve, located near pore field, within a stria replacing 5–6 vimines ( Figs 54, 56, 58, 60 ). A simple apical pore field at each pole, composed of round porelli. Outer surface of mantle and bands covered with small granules ( Figs 57, 59 ). Girdle composed of many open bands (5–11), with alternating orientation, with broad valvocopula and adjacent copulae. Valvocopula with rows of usually 2 (3) poroids on pars exterior; adjacent copulae with rows of 1 or 2 poroids, rarely three; pleurae (the last bands in any cingulum, following von Stosch 1975) with 1 row of poroids on pars exterior. Pars interior of all bands with one row of poroids, often elongate on valvocopula, elongate or round on other bands.

Type:— FAROE (FAERØERNE) ISLANDS (neotype designated here [holotype lost]: BM 21049!, “ Diatoma hyemale (Lyngb.) Ins. Faeroe. ” Figs 8–24 ).

Additional material examined: — FAROE (FAERØERNE) ISLANDS: “ Odontidium hyemale Kütz. Faröe. Lyngbye 1832. 49.”, (slide BM 17900!); “ Odontidium turgidulum Kütz. Faröe. Lyngbye. Hofm. Bang. 48.”, (slide BM 17908!); Eysturoy, south-east of Eiði, AErgisá, Faroe Islands, on mosses in a stream, (slide and material NMW.C.2015.001. Faroe.2014.5!). ITALY: Gesso della Valletta, Regional Natural Park of the Maritime Alps, Italy, rheocrene spring, (slide and material NMW.C.2012.021.2012. Italy.76!);. AUSTRIA: area of Schwarzmoos lakes near Kühtai, Stubaier Alps, Austria, spring, (slide and material NMW.C.2012.019.2012. Austria.13.sto.bryo.filalg!).

Observations:— The number of spines is highly variable from valve to valve comparisons. They sometimes form one to two rows of closely spaced interlinking spines. In some valves they are broken off but their locations are still visible; some valves have very few or no spines. In all valves with a single rimoportula it was positioned close to the pore field—one large valve had two rimoportulae near the pore field.

Nomenclature:— Kützing based Odontidium hyemale on “ Fragilaria hyemalis Lgb. ”, “Lgb.” is his abbreviation for Lyngbye (1819), he made reference to Fragilaria hyemalis in Agardh (1824: 7) but did not mention Conferva hyemalis Roth ( Kützing 1844: 44) . There are three possibilities for the source of the original description of Fragilaria hyemalis : Fragilaria hyemalis Lyngbye (1819: 185) , Fragilaria hyemalis C. Agardh (1824: 7) or Fragilaria hyemalis Kützing (1844: 44) . Agardh’s (1824: 7) Systema Algarum entry for “ Fr[agilaria] hyemalis ” attributes its origin to Lyngbye, making reference to an illustration from Lyngbye 1819 (“Lyngb. t. 63” = Lyngbye 1819: tab. 63, E 1–6, reproduced as Fig. 2 ; Agardh 1824: 8). Agardh further added that his concept of Fragilaria hyemalis “excl. syn. Roth”, which is a reference to Conferva hyemalis Roth ( Agardh 1824: 8) . Earlier, in his Dispositio Algarum Sveciae ( Agardh 1812: 28) and Synopsis Algarum Scandinaviae ( Agardh 1817: 78), Agardh had included and commented on Roth’s Conferva hyemalis as an independent species, so he was aware of the existence of Roth’s taxon. Much later, Crawford (1978: 247) wrote “… Agardh (1827) considers his M[elosira]. varians to be synonymous with C. hyemalis Roth, 1800a …”. In fact, Agardh referred to Fragilaria hyemalis Lyngbye which, as documented above, he did not consider to be synonymous with Conferva hyemalis Roth ( Agardh 1827: 628) . Crawford’s comparison of Conferva hyemalis with Melosira may account for the later synonym Lysigonium hyemale (Roth) Trevisan (1848: 96) , Lysigonium being a nom. rej. with respect to Melosira .

Because Kützing (1844: 44) did not include any reference to Conferva hyemalis but did cite Agardh, Silva & Hasle (2006: 200, 201) concluded that Kützing was following Agardh’s use of the name “ Fr[agilaria] hyemalis ”, which excluded Conferva hyemalis ( Kützing 1844: 44, an identical entry is found in Kützing 1849: 13) and concluded that “ Odontidium hyemale should be attributed directly to Kützing as the name of a new species that includes Fragilaria hyemalis sensu Lyngbye rather than as a combination of F. hyemalis Lyngb. ” ( Silva & Hasle 2006: 200, 201).

Yet Fragilaria hyemalis sensu Lyngbye (1819: 185) is based on Conferva hyemalis Roth (1800a: 506) as Lyngbye himself makes very clear ( Lyngbye 1819: 185, description reproduced as Fig. 4 ). As it happens, there is no indication in any of Agardh’s publications that he ever saw any specimens investigated by Roth. There was, however, a packet of material in C (C-A-94656!, see Fig. 3 ) labelled as C onferva hyemalis Roth containing specimens that Agardh had, presumably, studied. Examination of this material revealed two cells that may be identified as Melosira varians , otherwise there were no further specimens of Melosira nor of any species of Odontidium ; the species in abundance belong to Navicula Bory (1822: 128) and Surirella Turpin (1828: 362) .

There is no indication that Lyngbye saw any of Roth’s specimens either, but Kützing did make drawings from authentic Lyngbye material: “Unsere Abbildungen sind nach Lyngbye’schen Originalexemplaren aus der Sammlung des Hrn. Senator Binder angefertigt, welche von den Faröer-Inseln herrühren” ( Kützing 1844: 44, material is preserved as BM 17900!, of which there is also a packet of raw material stored in the herbarium. This material was designated as isotype material for Fragilaria hyemalis sensu Lyngbye in Williams 1985: 136). This suggests that the correct name for Fragilaria hyemalis should include reference to Conferva hyemalis Roth as its basionym and it would be Fragilaria hyemalis (Roth) Lyngbye (1819: 185) ; thus, as a species of Odontidium , the correct name is Odontidium hyemale (Roth) Kützing (1844: 44 , pl. 17, fig. 4), it also has Conferva hyemalis Roth as its basionym.

Roth (1800a) first described Conferva hyemalis from areas near Bremen, north Germany, but did not provide any drawings or references to previously published figures ( Roth 1800a: 506–507; species 22: “ In piscinis prope Schoenebeck non procul a Vegesack primo observavi, postea passim in fossis ” [= in fishponds near Schoenebeck not far from Vegesack seen first, later everywhere in ditches; Schönebeck is an area of Bremen –Vegesack]. Roth also gave a brief account of Conferva hyemalis in a second publication, Catalecta Botanica ( Roth 1800b: 205–206), where he listed Conferva hyemalis (species 20 on p. 205, 206) together with a short description, and a reference to “Tentamen Florae Germanicae, Tomus III, Pars 1, n. 22” ( Roth 1800b, in reference to Roth 1800a).

Roth’s collection was transferred to Berlin in 1925–1926. A large part of B was destroyed in 1943 (Håkansson in litt., Williams 1985, Regine Jahn, pers. comm.), which included Roth’s collections, so a suitable neotype is required. The most appropriate choice would be any of Lyngbye’s specimens from the Faroe Islands as his is the next available name and he provided several drawings of the species, its colony and filaments, but no clear view of the valves ( Lyngbye 1819: tab. 63, E 1–6, see Fig. 2 ). His specimens were collected from alpine rivers and other habitats from the Faroe Islands (“ Habitat in rivulis alpinis insularum Faeroesium …” Lyngbye 1819: 185, see Fig. 2 ). BM 21049! (from Roper’s collection) is slide no. 49 from Eulenstein’s exsiccata set ( Eulenstein 1867 –1869) and it is here designated as neotype of Odontidium hyemale . The upper label on BM 21049 reads: “Eul. Diat. spec . typ. no. 49. Diatoma hiemale (Lyngb.) Ins. Faeroe. In aqua dulci.”. Its lower label reads: “Spec . originale. Lyngb. Tent. Pag. 185. t. 63.”. The upper label of the slide was subsequently annotated by Robert Ross who corrected the specific epithet to “ hyemale ” and added in a further note to indicate the possible use of the name Odontidium should it have been given priority ( Fig. 7 ). It should be noted that Eulenstein produced two editions of the first 100 slides in his exsiccata set, the first issued in 1867, the second in 1869 (for a discussion of these sets of slides see Quarterly Journal of Microscopical Science 1867: 64). Some of the slides differ from edition to edition. In the first edition slide no. 49 was made from material that came from France and is not Lyngbye’s material: BM 21049a, “ Vogesus , Gallia . In aqua dulci ” = Diatoma maximum in Williams 1985: 145. Thus, it is expected that there are duplicate slides of this neotype elsewhere (there is one additional no. 49 “ Ins . Faeroe ” in BM as an unnumbered slide, and one additional no. 49 “ Vogesus , Gallia ” in BM also as an unnumbered slide).

Slide BM 17900 is labelled “ Odontidium hyemale Kütz. Faröe Lyngbye. 1832. 49.” ( Figs 7 , 25-31 ), the specimens resemble smaller valves of the neotype. As noted above, Kützing’s line drawings of Odontidium hyemale are based on Lyngbye’s material (“Originalexemplare”, Kützing 1844: pl. 17, fig. 4, reproduced here as Fig. 5 ). Only the larger valves of O. hyemale are depicted. Smaller valves drawn by Kützing (1844: pl. 17, fig. I), resembling the specimens in our Figs 25–31 , were recorded by Kützing as Odontidium mesodon ( Kützing 1844: 44) .

BM 17908! (Coll. Kützing) is labelled “ Odontidium turgidulum Kütz. Faröe. Lyngbye. Hofm. Bang. 48.” ( Fig. 7 ) and contains a small number of smaller valves belonging to other species, and some small valves of Odontidium hyemale as seen on slide BM 17900. Kützing’s drawings ( Kützing 1844: pl. 17, fig. II) depict valves smaller than those of O. hyemale ( Kützing 1844: pl. 17, fig. IV, 1–4), and longer, slimmer valves as compared to O. mesodon ( Kützing 1844: pl. 17, fig. I). Given the few valves present on BM 17908 it is not clear which valve features he refers to. Kützing acknowledges Ehrenberg’s Fragilaria turgidula Ehrenberg (1838: 204) as a possible earlier example, specimens that require re-examination if this is the basionym of Kützing’s O. turgidulum ( Fig. 6 ).

Grunow (1862) described Odontidium hiemale var. genuinum Grunow (1862: 356) but provided no drawings. He listed, among other names, Fragilaria hiemalis and Odontidium hiemale as synonyms (for the latter he referred to the accounts in Kützing 1844: pl. 17, fig. 4 and Smith 1856: pl. 34, fig. 289). He described the valve margins as long, elliptic-lanceolate with 6–12 transapical ribs. He mentioned that O. hiemale var. genuinum occurs without the other taxa he regards as varieties, Odontidium hiemale var. turgidulum (Ehrenb.) Grunow (= O. turgidulum Kützing ) and Odontidium hiemale var. mesodon (Ehrenb.) Grunow (= Odontidium mesodon Kützing ), in very cold waters at higher elevations in the Alps or in the north, but is also often associated with the latter.

Heiberg used the name “ Diatoma hyemale (Lyngbye) ” and referred to “ Fragilaria hyemalis Lyng. ” (= Lyngbye 1819) as the older synonym ( Heiberg 1863: 58), based on “ Conferva hyemalis Roth ” ( Heiberg 1863: 16); O. hyemale Kützing and O. mesodon Kützing are both listed as synonyms; Heiberg refers to Smith’s Synopsis ( Smith 1856: pl. 34, figs 288, 289) and cited Smith’s remarks on the “doubtfully distinct” nature of the two ( Smith 1856: 16). There are no illustrations in Heiberg.

Figures

FIGURES 8–31. Odontidium hyemale, LM, valve views. 8–24. BM 21049, “Diatoma hiemale/hyemale, Eul. Diat. spec. typ. No. 49”, Faroe Islands, neotype. 25–31. BM 17900, “Odontidium hyemale Kütz., Lyngbye 1832. 49.”, Faroe Islands. Scale bar = 10 μm.

FIGURES 32–35. Odontidium hyemale, LM. 32. Girdle view and valve view. 33–35. Girdle views. 32–33. BM 21049, “Diatoma hiemale/ hyemale, Eul. Diat. spec. typ. No. 49”, Faroe Islands, neotype. 34–35. BM 17900, “Odontidium hyemale Kütz., Lyngbye 1832. 49”, Faroe Islands. Scale bar = 10 μm.

FIGURES 36–43. Odontidium hyemale, LM, valve views. NMW.C.2015.001.Faroe.2014.5. Eysturoy, Faroe Islands, stream. Scale bar = 10 μm.

FIGURES 44–52. Odontidium hyemale, LM, valve views. NMW.C.2012.021.2012.Italy.76. Gesso della Valletta, Regional Natural Park of the Maritime Alps, Italy, spring. Scale bar = 10 μm.

FIGURES 53–60. Odontidium hyemale, SEM. 53–55, 57–60. NMW.C.2015.001.Faroe.2014.5. Eysturoy, Faroe Islands, stream. 56. NMW.C.2012.019.2012.Austria.13.sto.bryo.filalg. Area of Schwarzmoos lakes near Kühtai, Stubaier Alps, Austria, spring. 53. View of whole valve and girdle bands. 54, 56. View of whole valve. 55, 57. View of whole frustules in girdle view. 58. Internal view of whole valve. 59. Details of valve mantle and girdle band. 60. Internal view of apical pore field and rimoportula. Scale bars = 2 μm (Fig. 57), 1 μm (Figs 53–56, 58–60).

FIGURES 2–5. Reproductions of published drawings and type material. 2. Drawings of “Fr[agilaria] hyemalis” from Lyngbye (1819: tab. 63 E. 1.Algae tuft. 2, 4–6. Enlarged filaments. 3. Dried enlarged filaments). 3. Packet and dried material from C (C-A-94656) labelled as Conferva hyemalis Roth, possibly original material and seen by Agardh. 4. Reproduction of Lyngbye’s description of “Fragilaria hyemalis” with reference to Conferva hyemalis Roth (Lyngbye 1819: 185). 5. Reproduction of drawings from Kützing (1844: pl. 17, IV, 1–4) of Odontidium hyemale, based on Lyngbye’s material (“Originalexemplare”, collection Senator Binder, from the Faröe Islands, Kützing 1844: 44).

FIGURES 6–7. Reproductions of published drawings and type material. 6. Reproduction of Ehrenberg’s drawings of Fragilaria turgidula Ehrenberg (1838: 204). 7. Type slides for Odontidium hyemale Kützing (Faröe, BM 17900), Odontidium turgidulum Kützing (Faröe, BM 17908), and Diatoma ‘hiemale/hyemale’ (Eulenstein, Diat. spec. typ. No. 49, BM 21049, neotype).