Navicula cryptocephala 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 : 30-31

Navicula cryptocephala Kützing 1844 ( Figs 87–96 , 165 )

Frustules narrowly lanceolate in valve view, 25.1–30.0 µm long, 4.8–5.5 µm wide, with subrostrate apices, biraphid. Living cells containing 2 platelike parietal chloroplasts, one each side of the cell ( Figs 87, 88 ). Areolae elongate and longitudinally aligned; striae thus uniseriate and lineolate, (17–) 18–20 in 10 µm, radiate around central area ( Fig. 95 ) to transverse at poles ( Fig. 96 ). Central area widened, with (4–)5 shortened striae ( Fig. 95 ). Raphe straight (may be slightly undulate near central termini), with slightly expanded central termini and polar termini hook-shaped over the valve apices ( Fig. 96 ).

Reference: — Krammer & Lange-Bertalot (1986) p. 102–103, fig. 31: 8–14; Cox (1995) p. 108, figs 23–27, 50–54, 73.

Specimens examined: —CHR618415! (cleaned frustules from cultured strain LCR-S:32:1).

Distribution: —The species is regarded as widespread and cosmopolitan. However, Poulíčková & Mann (2006) regarded biogeographic conclusions based on use of the name N. cryptocephala as “almost meaningless”, owing to the variation in reproductive features that they observed between populations in this complex. Based on their results, it seems likely that such variation would also occur within our N. cryptocephala clade. Organisms ascribed to this name are well known from New Zealand, and they appeared in cultures from the Styx from site 10.

Molecular data: —Only a short 18S sequence of 344 bp was obtained for this strain. The closest matches to this fragment were N. cryptocephala strains AT-114Gel08c, AT-176Gel05, UTEX FD109, and Navicula sp. GGM-2004a (p-distances = 0.003). The first of these strains is from Germany, the second from Minnesota, and the third is from an unknown site. They formed a robust clade with the Styx strain in both Bayesian and MPB analyses ( Fig. 165 ). The overall dataset was 1795 bp long, including 148 variable sites (92 parsimony informative, 14 occurring within the Styx fragment). The model selected and implemented in the Bayesian analysis was TN93+G+I.

Observations: —This strain may be easily distinguished from N. veneta Kützing 1844 (which is also found in the Styx River) by its larger size, strongly radiate striae around the central area with more shortened striae, absence of convergent striae near valve apices, and absence of hooked central raphe endings. The two are also clearly distinguished in molecular analyses. N. cryptocephala LCR-S:32:1 is a close match to published illustrations of this species ( Krammer & Lange-Bertalot 1986), and the features that distinguish it from the Styx strain of N. veneta are constant throughout. There are some subtle differences between our material and Kützing’s type material as illustrated and described by Cox (1995): the Styx strain has more subrostrate ends, usually with a detectable point of inflection between the apices and the main part of the valve ( Figs 89, 96 , which is similar in some of Kützing’s illustrations), and has slightly denser striae (cf. 16 in 10 µm in Kützing’s collection).

Navicula trivialis Lange-Bertalot (1980: 31) has been shown to have a very close relationship with N. cryptocephala ( Poulíčková et al. 2010) . One strain studied by these authors formed a clade with N. trivialis ; this strain was indistinguishable from their Clade I ( N. cryptocephala only) by valve length, breadth, or stria density. However, 18S data are not available for N. trivialis , and the strains used by Poulíčková et al. 2010 differed, so it is not possible to say whether our strain of N. cryptocephala belongs to their Clade I or II (or indeed, a further clade).

Figures

FIGURE 165. Bayesian phylogenies of Navicula, Sellaphora, and Eolimna strains from the Styx River, for 18S rDNA (left) and rbcL (right). 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 87–96. Navicula cryptocephala strain LCR-S:32:1, CHR618415. Scales: Fig. 87, 10 µm (use for Figs 87, 88); Fig. 89, 5 µm (use for Figs 89–93); all others 1 µm. Figs 87, 88. Live cells viewed by LM, showing paired parietal plastids close against valve margins. Figs 89–93. Cleaned frustules viewed by LM (girdle view in Fig. 93, showing curved valve faces). Figs 94–96. SEM of cleaned frustules. Fig. 94, whole frustule, showing elongated, longitudinally aligned areolae; Fig. 95, widened central area showing central raphe termini; Fig. 96, valve apex showing hooked raphe terminus and pattern of areolae.