Stenopterobia colombiana Sala, Guerrero, Nunez Avellaneda & Kociolek 2021, sp. nov.

Stenopterobia colombiana Sala, Guerrero, Núñez Avellaneda & Kociolek sp. nov. (Figs 20–37)

Description: —Valves are sigmoid, with both sides S-shaped, parallel at more than 1/3 of the valve, usually slightly constricted at mid-valve, and narrow protracted ends (Figs 20–33). Externally, the valve face is slightly transapically undulated with a wide, sunken axial area; in some specimens there are scattered shallow depressions and a faint, elevated axial rib in the median line (Fig. 34, arrow). Transapical costae occupy about 2/3 of the valve face and towards the margin they are fused to the raphe canal; some of them, alone or grouped in pairs, become slightly slender before the quite inconspicuous fenestrae (Fig. 34). Other costae, arranged in pairs or groups of three, have the same breadth lengthwise. Costae alternate with narrower, biseriate striae with small round areolae arranged in a quincunx pattern (Fig. 34). The raphe canal is located on a barely developed wing (Fig. 34). The valve mantle has alternate costae and striae and its margin is occupied by a thick rim. Internally, the axial area is smooth, wide at the valve center (Fig. 35) and tapering to the ends (Fig. 36). The portulae are slightly elongated and irregularly spaced; the areola openings are not rimmed (Figs 35–36). Raphe ends are simple and discontinuous (Fig. 37).

Dimensions (n=50): length: 73–99 µm; width: 3.5–5.5 µm; portulae: 93–108 in 100 µm; striae: 34–40 in 10 µm; areolae: 96–120 in 10 µm.

Type: — COLOMBIA. Caquetá River Basin: Orteguaza River, flooded forest creek, 1º 38’ 1.46” N, 75º 28’ 45.98” W, M. Núñez Avellaneda, 1 st November 2003 (holotype slide LPC 5610-2! finder: G/33–2–4, here illustrated as Fig. 29).

Etymology: —The species epithet refers to the country where the new species was collected.

Comments: — Table 2 summarizes the main morphometric parameters of the new species and those of previously described sigmoid Stenopterobia species. Although S. colombiana coexists with S. caquetaensis, they are distinguished by their valve dimensions (valve length of S. colombiana is generally shorter) as well as by their portula, stria and areola densities. Moreover, they differ in valve outline; both species have attenuated apices but in S. colombiana valve ends are more abruptly protracted and thus more clearly differentiated form the main body of the valve, which in addition is slightly constricted in the middle. These two species also differ in valve structure, particularly in the type of striae (always biseriate in this species while bi-, tri-, rarely tetraseriate in S. caquetaensis), the relative width of striae and costae, and the width of the axial area.

The studied materials can be clearly differentiated from S. curvula by their valve apices much more protracted, wider axial area, striae constantly biseriate, generally lower dimensions and a higher portula density.

Although they have a similar cell size, S. densestriata differs from the new taxon in its lower stria, areola and portula densities (26–30 in 10 µm, 60 in 10 µm and 40–60 in 100 µm, respectively) and in its valve face morphology. Even though the specimens described as S. cf. densestriata in Ruck & Kociolek (2004) resemble S. colombiana in valve outline, the former differs in the presence of a very narrow axial area and bi- or triseriate striae (Ruck & Kociolek 2004, pl. 59, fig. 7). A further difference is that in our specimens all costae are fused to the raphe canal whereas in S. cf. densestriata those between alar canals proceed below the raphe canal (see pl. 59, fig. 7 in Ruck & Kociolek 2004).

As mentioned above for S. caquetaensis, comparison of S. colombiana with S. planctonica is difficult. Both species are remarkably similar in valve dimensions and outline (compare our LM Fig. 8 with fig. 219: 3 in Metzeltin & Lange-Bertalot 1998) but S. planctonica has lower portula and stria densities (60–90 in 100 µm and 29–31 in 10 µm, respectively) and the areolation pattern is different (striae with 3–4 rows of areolae). No SEM images of the valve exterior of the type specimens are available, therefore further comparisons are not possible.

Stenopterobia pelagica is similar to S. colombiana in some valve morphometric data such as width and stria density and in the occurrence of narrow valve apices. With regard to the density of portulae, we observed differences between the type material from the Célebes Island (Simonsen 1987) and specimens from South America (Metzeltin & Lange-Bertalot 1998, Oliveira et al. 2012). Specimens from Célebes have 60–80 portulae in 100 µm, while specimens from South America have 120 in 10 μm (Metzeltin & Lange-Bertalot 1998) and 60 portulae in 100 µm (Oliveira et al. 2012); however, none of them coincide with the Amazonian materials. Siver & Camfield (2007) studied Hustedt’s materials, identified as S. pelagica, with SEM from Lake Wawontoa, a site near the type locality but images provided do not allow the measurement of the number of portulae. Although these specimens have biseriate striae like those found in our species, they differ in having a slender axial area and interestriae narrower than the striae.