Basionym: Navicula cocconeiformis Gregory ex Greville (1855: 256, fig. 9: 6)
Valves lanceolate to elliptic-lanceolate in smaller specimens with rostrate to broadly rounded apices. Valve length 12–25 (36) µm, width 7–11 (15) µm. Striae curved, radiate at center, to strongly radiate at apices, 25–30 in 10 µm; around central area striae alternate long and short. Striae uniseriate with fine rounded, to moderately elliptical areolae, which continue onto valve mantle. Internally, areolae positioned between thickened costae and covered with a raised hymen. Axial area narrow, linear to weakly lanceolate along valve length; central area not defined. Raphe distinct and central on valve. Externally, proximal raphe features tear-drop like; distal raphe features terminate on valve face, bent strongly or weakly in opposite directions at apices. Internally, proximal and distal raphe features terminate with helictoglossa. Copulae open bands, with two rows of pores.
Type:— Elchies and various places in Banffshire, and latterly from Loch Leven (Gregory 1855: 256) (BM 566, “Braemar, Balfour, 8.54, no. 120” as Navicula nugax). “Elgin”, lectotype slide BM 689, lectotype designated here, BM 689.
Observations:— Two morphotypes have been identified based on valve shape, distal raphe fissures and the presence of terminal areolae. Morphotype 1 has rounded to weakly rostrate apices with strongly deflected terminal fissures and no areolae in the terminal (apex) hyaline area (Figs 17–31, 136–139). Morphotype 2 (Figs 2–16, 132–135) has rostrate apices with weakly to strongly deflected distal raphe fissures and at least one series of areolae extending down the mantle at the apex (Figs 132, 133, 140). The specimens (morphotype 1), depicted as C. cocconeiformis resemble N. cocconeiformis (Greville 1855, fig. 9: 6) with a lanceolate valve outline and rostrate apices (Figs 28, 29). Slides in the Gregory collection are poorly made, labeled and indexed (Williams pers. com.). Greville’s slides are the next best thing as they were prepared from Gregory’s material. The slide clearly labelled in the Greville collection as N. cocconeiformis was BM 689 from Elgin. The Elgin locality was not listed in Greville’s original 1855 publication, but in 1856 Gregory (Notice of some new species of British freshwater Diatomaceae) lists the Elgin locality first in the “new species description and in the line drawing Gregory presents Navicula cocconeiformis under the section title “New species, now first named”. With this we feel confident that both Greville’s and Gregory’s concept of the species is depicted by specimens from Elgin and here designated as the lectotype.
Morphotype 2 matches the line drawing of N. cocconeiformis (Gregory 1856, p. 6; pl. 1: 22) which in the figure legend states “new species, now first named”. Although it is tempting to create a new species from morphotype 2, without seeing SEM images of Gregory’s original syntype materials we cannot effectively discern which morphological form is C. cocconeiformis sensu stricto as intended by Gregory. DNA studies may assist in evaluating differences in these morphotypes. C. cocconeiformis is designated as the type species for the genus Cavinula (Mann & Stickle in Round et al. 1990) and morphotype 1, as presented, is considered representative of C. cocconeiformis sensu stricto.
Krammer & Lange-Bertalot (1986, fig. 59: 2–5) depict valves representing morphotype 1 with a more rhombic outline and defined by the following valve features: L=12–40 µm, B=7–15 µm, and 24–36 striae in 10 µm. Morphotype 2 resembles the specimens presented in Siver et al. (2005, fig. 38: 1–3). They reported the following features for the population observed: valve length 21–26 µm, width 11–12 µm and stria density 24–28 in 10 µm. Foged (1971) documented C. cocconeiformis as a common taxon occurring in a small deep lake on the Northern Slope (Alaska), but states that “none of the presented illustrations represent the typical form” (Foged, 1971, fig. 15: 16, 18, 19). One of the specimens illustrated (Foged, 1971, fig. 15: 16) would match morphotype 2, while the other specimens are C. cocconeiformis f. elliptica (Foged, 1971, fig. 15: 18) and C. vincentii (Foged, 1971, fig. 15: 19). Morphotype 2 here noted, also resembles C. pusio (Cleve) Lange-Bertalot, but is distinguished by the deflected distal raphe ends and areolae extending down the mantle at the apex. Antoniades et al. (2008) reported specimens from the Arctic Archipelago for morphotype 1 (fig. 50: 7, 8) and morphotype 2 (fig. 50:10–12) with valve lengths 12–24 µm, widths 6–9.5 µm and stria densities 22–26 in 10 µm. C. cocconeiformis sensu lato can be distinguished from other Cavinula taxa by valve outline, the absence of a clearly defined central area, the strongly curved striae between valve center and apex, stria density and the deflected distal raphe endings in opposite direction at both valve apices.
Distribution: —During our observations we confirmed the presence of C. cocconeiformis from White Pond (Ellesmere Island, Nunavut Territory), Maskinonge, St-Maurice, Lac aux Araignees and Madawaska (province of Québec), Lake Contwoyto and Lac De Gras (Northwest Territories, Singleton Lake, Clear Lake, Little Clear Lake, Sanschambre Lake, Ashigami Lake, Pedro Lake, Wawashkesh Lake, Crosson Lake, Laundrie Lake and Moot Lake (province of Ontario). In the CANA collection this taxon was identified from another 174 Canadian localities. C. cocconeiformis was also confirmed in the samples from Big Moose Lake, Pine Pond, Moriah Parch Pond and Santa Clara Green Pond (Adirondack Park, the state of New York, USA), and in 16 samples from ponds on Cape Cod. C. cocconeiformis was also tentatively identified in 316 + waterways across the United States (ANSP collection); however, voucher photomicrographs indicate that some of these identifications represent more than one taxon. We have verified the biogeographic distribution of C. cocconeiformis from Colorado to Maryland, extending north to northern Ellesmere Island and Greenland. Average weighted means for pH vary from 6.0 (northeastern United States) to 7.1 on the high Arctic Islands. The estimated weighted means for total phosphorus varied from 8.0 to 14.4 µg/L with a large error estimate. Conductivity values were low (<200 µS/cm) and dissolved organic carbon values varied from 1.1–2.6 mg /L. Although commonly observed, this taxon was never abundant with relative occurrences of <5%.
Distribution Records:— Foged (1971, fig. 15: 16, 18, 19, Alaska), Foged (1974, fig. 17: 1–4, Iceland), Foged (1977, fig. 26: 14, Ireland), Krammer & Lange-Bertalot (1986: 158, fig. 59: 2–4, Europe), Watanabe (2005, fig. II B3- 18: 1–3, Japan), Antoniades et al. (2008: 47, fig. 107: 5, Arctic Archipelago), Potapova (2014, figs 211, 212, Bering Island, Kamchatka Peninsula, Russia).