Cyclotella meneghiniana Kützing 1844

Novis, Phil M., Braidwood, Jasmine & Kilroy, Cathy, 2012, Small diatoms (Bacillariophyta) in cultures from the Styx River, New Zealand, including descriptions of three new species, Phytotaxa 64 (1), pp. 11-45 : 19-20

Cyclotella meneghiniana Kützing 1844 ( Figs 2–12 , 160 )

Frustules 11.7–12.8 µm wide in valve view, 5.3–9.6 µm in girdle view. Chloroplasts of living cells brown, parietal, plate-like; central nucleus and/or vacuole occasionally visible. Valves areolated peripherally, with 26–32 radial costae. Central area 5.4–7.6 µm wide, not areolated, smooth internally, colliculate externally, tangentially undulate. One marginal fultoportula usually present on each costa ( Figs 10, 11 ), with a long spine inserted above the outer opening and a shorter spine at each of the three remaining sides. Central fultoportulae offset with respect to cell centre, occurring in the raised part of the undulate central area, 1–2 per valve. Rimoportulae approximately 0.2 µm wide, 0.6 µm long, one per valve ( Fig. 11 ).

References: — Beszteri et al. 2005a, p. 92, figs 1, 3–10; Kützing 1844, p. 50, fig 68; Skvortzov 1938, p.412–413, pl. 1 fig. 38.

Specimens examined: —CHR618421! (cleaned frustules made from culture LCR-S:46:3).

Distribution: — C. meneghiniana is regarded as cosmopolitan and is well known from New Zealand (Cassie Cooper in press). Within the Styx River, the species appeared in cultures from downstream sites (9 and 11).

Molecular data: —The closest match to the Styx strain was strain TI1 (p-distance 0.002), isolated from

Lake Titicaca, Peru. Other close matches were strains LS03-01 from Lake Superior, MI, USA and L1263 from Stump Lake, ND, USA (both 0.003). The Styx strain grouped with all other C. meneghiniana strains in the database, forming a robust clade in both Bayesian and MPB analyses that also included C. gamma Sovereign 1963 and C. cryptica Reimann et al. 1963 ( Fig. 160 ). The full dataset included 1473 bp with 270 variable sites (169 parsimony-informative, 82 in our 581-bp fragment). The substitution model chosen using BIC and implemented in the Bayesian analysis was GTR+G+I. The Bayesian analysis did not support monophyly of the genus Cyclotella Brébisson 1838 , as previously noted ( Alverson et al. 2007).

Observations: —The very small difference in rbc L sequence with the Lake Titicaca strain suggests that the Styx C. meneghiniana is not range restricted. The species has long been considered cosmopolitan, although significant cryptic diversity is known to exist ( Beszteri et al. 2005 b, 2007). However, the component clades have not been subject to any nomenclatural division to date, and our material agrees with isolates placed in C. meneghiniana at the morphological and molecular levels. An entity was previously described from an aquarium in Napier, NZ, as Cyclotella meneghiniana var. novaezelandiae Skvortzov (1938: 412) . The valve face of this variety was 8.5 µm diameter, whereas the Styx samples are> 10 µm, and the radiating costae extend much more closely to the centre of the valve face than is the case for the type variety (and the Styx specimens). A comparison of Kützing’s (1844) and Skvortzov’s (1938) original illustrations makes this latter feature very obvious. We note that the width of the central area is regarded as a diacritical measurement in these morphotypes ( Beszteri et al. 2005a), so there seems no possibility that our specimens correspond to the earlier “ New Zealand ” variety.

Subclass Bacillariophycidae

Superorder Bacillarianae

Order Achnanthales

Family Achnanthidiaceae

Figures

FIGURE 160. Bayesian phylogenies of the Cyclotella strain from the Styx River, for rbcL. Support numbers above branches are Bayesian posterior probabilities / Maximum Parsimony bootstrap values. Scale bar = 0.1 subsitutions/site. See text for substitution models and description of data sets.

FIGURE 2–12. Cyclotella meneghiniana, strain LCR-S:46:3, CHR618421. Scales: Fig. 2, 10 µm; Fig. 3, 5 µm (use in Figs 3–8); all others, 1 µm. Fig. 2. Live cells viewed by LM in valve, oblique, and girdle views (marginal spines are visible on oblique cell). Figs 3–8. Cleaned frustules viewed by LM, showing size variation and length of costae (Figs 7, 8, through-focus image pair of cells in girdle view). Figs 9–12. Cleaned frustules viewed by SEM. Fig. 9, whole frustule in slightly oblique valve view; Fig. 10, ornamentation in the presence and absence of marginal fultoportulae; Fig. 11, valve interior showing central fultoportulae and rimoportula; Fig. 12, higher magnification view showing structure of central fultoportulae.