Nitella caespitosa C.C. Jao & Y.Y. Lee

Romanov, Roman E., Nikulin, Vyacheslav Yu., Yeo, Regina S. W., Ho, Boon-Chuan & Gontcharov, Andrey A., 2020, Nitella singaporensis (Charophyceae, Charales): new species and implications for the taxonomy of the genus Nitella, Phytotaxa 438 (2), pp. 80-94 : 88-90

Nitella caespitosa C.C. Jao & Y.Y. Lee in Y.Y. Lee, Acta Phytotax. Sin. 19(2): 168 (1981)

≡ Nitella glomerata auct . non (Desv. in Loiseleur 1810: 135) Chevalier (1827: 124): C.C. Jao & Y.Y. Lee, Acta Phytotax. Sin. 12(3): 354, Pl. 71: 13–20. 1974 nom. illeg. Type (translated from Jao & Lee (1974) in Chinese):— China, Wuhan, northern foot of Moshan, growing in the shallow water pond at the East Lake, 18 IX 1963, Y.Y. Lee HP6831 (holotype: HBI n.v.).

Below we have noted some differences between Nitella singaporensis with the diagnosis and illustrations of N. caespitosa provided in Jao & Lee (1974, as N. glomerata ). Even though we have not examined the type specimen of N. caespitosa (holotype: HBI), the protologue ( Jao & Lee 1974) differs from N. singaporensis in:

1. Shorter relative length of the first segment of a branchlet, ca. 2/5 or slightly shorter than ½ of a branchlet according to the drawing of branchlets from two whorls only ( Jao & Lee 1974: pl. 71: 13, 16), i.e. nearly overlapping with N. singaporensis .

2. Longer branchlet rays, overlapping with N. singaporensis in case of primary ray only. However, these differences seem to be less prominent if dimensions of comparable parts of the branchlets are taken from the drawings with a scale reproduced by Han & Li (1994: fig. 46). The illustrations in Jao & Lee (1974) are published without a scale.

3. More branchlets in a whorl, 7–9 according to the description, 8 or 9 according to the drawings in Jao & Lee (1974: pl. 71: 13, 16). Therefore, these values are overlapping with N. singaporensis having 6 or 8 branchlets in each whorl.

4. Less variable number of lateral secondary rays of 5 or 6 according to the protologue, but 4 to 6 according to the drawings in Jao & Lee (1974: pl. 71: 13, 16), thus overlapping with N. singaporensis with 4–7 secondary rays.

5. Fewer lateral tertiary rays (4 or 5) at the apex of central secondary rays according to the drawing of a single whorl in Jao & Lee (1974: pl. 71: 13, 16), overlapping with N. singaporensis (5 or 6 lateral tertiary rays).

6. Apex of penultimate cell of a dactyl is described as rounded according to the description, but narrowing toward the apex with subtle difference in width between it and base of end cell according to the drawing in Jao & Lee (1974: pl. 71: 14). Our specimens of N. singaporensis resemble these drawings more closely. The slightly narrower end cells of dactyls in N. singaporensis are likely an artefact from rewetting dried specimens.

7. Less spreading dactyls according to the drawings in Jao & Lee (1974: pl. 71: 13, 16–19); in N. singaporensis the angle between dactyls is slightly bigger ( Figs. 9, 12–17 ).

8. Shortly stipitated oogonia and antheridia according to the protologue where a very short stipe of oogonium was illustrated by Jao & Lee (1974: pl. 71: 20). The short stipes have been found in N. singaporensis for antheridia only. The dissection was not possible for pressed specimens which are modest in amount, but gametangial position and structure are similar to the illustration in Jao & Lee (1974).

9. Smaller oogonia, oospores and antheridia dimensions, i.e. 224–239 × 159–200, 150–190 × 130–160, 198–205 μm, respectively, according to the protologue. These values are not overlapping with ones in N. singaporensis , which are bigger—393–429 × 329–340, 275–300 × 239–243, 229–234 μm. This is one of the most important difference with N. caespitosa .

10. Difference in oospore surface ornamentation, described as granulated (translated from Latin) or having granular protrusions (translated from Chinese) in the protologue ( Jao & Lee 1974) or granulated as elsewhere ( Ling et al. 2000), but never illustrated.

Therefore, the most significant differences between Nitella caespitosa and N. singaporensis are the differences in the size of gametangia and oospores, as well as the appearance of oospore surface in light microscope. The gametangia and oospores of the type of Nitella caespitosa need to be re-examined critically to confirm their dimensions and oospore surface ornamentation pattern. This is essential because the oospore ornamentation is a taxonomic trait of primary significance for Nitella classification and species differentiation ( Sakayama 2008; Casanova 2009). The type of surface ornamentation pattern found in Nitella singaporensis has not been reported yet in Nitella , although we have looked through numerous published SEM studies of oospores of Nitella species from nearly all around the world (e.g. Leitch et al. 1990; Sakayama et al. 2002, 2005; Casanova 2009; etc.).

The important issue arises from the presence of cells subtending from the base of the branchlets, which were found in Nitella singaporensis . It seems to be a common trait in Nitella and Tolypella (A. Braun) A. Braun (1857: 338) , both described as having no stipulodes. These cells can be easily spotted with good photos of stem nodes, even from unrelated species from different sections of the both genera ( Table 1). The similar arrangement of stem nodal cells had been found for N. gracilis collected in Sicily ( Romanov et al. 2019), where protruding cell surrounded the base of a branchlet ( Fig. 31 ).

The presence of cells below branchlet base has been observed in Nitella fuana Lan & Qing in Chen et al. (2018: 130, 131: fig. 1; the name ‘ N. fuiana ’ in the protologue has been corrected by Guiry & Guiry (2019) according to Turland et al. 2018: art. 60.8 (c)), but obviously not similar with N. caespitosa in all other traits. Nitella fuana was recently described from China with cells at the base of branchlets as one of the main morphological traits differentiating it from the similar N. mirabilis Nordstedt ex Groves (1924: 364 , pl. 35) ( Chen et al. 2018). These structures are erroneously referred to as stipulodes, although it is clearly visible from photos that they are extruding from the base of the branchlets (cf. Chen et al. 2017: Figs. 1b, c, and 2c, d ; Chen et al. 2018: Fig. 2d, e ). Therefore, we are suggesting that Nitella should remain to be characterised as stipulode-less genus of charophytes.

This calls for a reassessment of the taxonomic boundary between Nitella fuana and N. mirabilis as both species are sharing many common characteristics. Nitella fuana differs from N. mirabilis mainly by the presence of cells below branchlet base, but as explained above, the presence of this trait seems to be non-crucial. The other differentiating traits suggested by the authors include presence of a cell longer than others in upper row of oogonium coronula, formation of gametangia both at branchlets and their whorl bases, producing of 1–3 oogonia at the branchlet node in comparison with 2 or 3 in case of N. mirabilis , as well as differences in descriptions of oospores surface studied in light microscope ( Chen et al. 2018). Unfortunately, genetic data for N. mirabilis were not taken into account. Trait differences, such as cell length in coronula upper row, oospore surface pattern, and formation of stipulode-like cells, need to be confirmed with the reassessment of the type specimen of N. mirabilis . Critical comparison of existing data for N. mirabilis is challenging because of the short protologue ( Groves 1924) and the lack of details in later work ( Wood & Imahori 1964, 1965), although the similarity in oospore surface ornamentation in photo of N. fuana and drawing of N. mirabilis can be noted (cf. Groves 1924: pl. 35, 7; Chen et al. 2018: fig. 1f).

The difference in oogonia number between Nitella fuana and N. mirabilis cannot be accepted. The protologue of N. mirabilis states that “oogonia usually clustered 2–3 together, a few solitary” ( Groves 1924: 364), which has been confirmed by Wood & Imahori (1965: 391): “gametangia solitary, geminate or aggregated”. According to the protologue, N. fuana has 1 or 2 or many stipitate gametangia aggregated at the base of whorl and on branchlet nodes. Therefore, N. mirabilis has 1–3 or many oogonia in each node as in N. fuana . The gametangia arrangement is an important characteristic for species delineation in Nitella . Unfortunately, N. mirabilis has a short protologue without detailed description of its gametangia arrangement. According to the original drawing, its gametangia are seemingly forming not only at branchlet nodes but also sometimes at the base of branchlet whorls ( Groves 1924: pl. 35: 1, 2). The later reinvestigation of its syntypes by Wood & Imahori (1964, 1965) has confirmed this and clearly illustrated by Imahori ( Wood & Imahori 1964: icon 185). Moreover, the drawing of N. mirabilis var. libera F.S. Han & H.L. Fu in Fu & Han (1990: 76, 77, fig. 1) described from China clearly illustrates abundant oogonia at the base of the whorl ( Fu & Han 1990: fig. 1, 1). Therefore, we question the recognition of N. fuana from N. mirabilis by morphology. We suggest a search for other morphological traits suitable for critically separating these two species, otherwise their conspecifity is indicated.

According to the results of our rbc L-based phylogeny, Nitella comptonii and N. vieillardii are the closest relatives of N. singaporensis ( Fig. 38 ). Therefore N. singaporensis clearly belongs to Nitella subg. Tieffallenia sect. Gioallenia ( Wood & Imahori 1965) according to its morphology and phylogenetic affinity. Comparison of the most similar species with their phenotypic traits has revealed some essential morphological differences.

Nitella comptonii differs by the 3–4-furcated branchlets, clearly confluent end cell of a branchlet, and finely granulate oospore surface according to the protologue ( Groves 1922) or finely granulate under LM and fibrous or spongy with irregularly arranged granules under SEM according to Sakayama et al. (2006). The species has central secondary ray, but does not has central tertiary ray ( Groves 1922: pl. 5, 2).

Nitella vieillardii differs mainly by the 2–3-celled dactyls and reticulated surface of the oospore. In addition, its plants may have submucronate end cells, central secondary branchlet rays, and fertile basal branchlet nodes ( Wood & Imahori 1965), which is interpreted as an absence of central secondary ray in the last case and sterile in the previous ones or producing only oogonium at the basal node of a branchlet in case of central secondary ray presence. The plants with strictly 2-celled dactyls referred to the species has irregular or imperfectly reticulated oospore surface under LM and SEM having a strongly spongy texture ( Sakayama et al. 2005). This pattern is in obvious contrast with N. singaporensis .

In Chinese charophyte flora Nitella caespitosa (as N. glomerata ), most similar with N. singaporensis , has been compared with N. flagellifera Groves & Allen (1927: 337) , as well as with N. gracillima Allen (1898: 76–77) and N. dictyosperma Groves & Groves (1898: 324–325 , pl. 19) ( Han & Li 1994). Nitella flagellifera and N. gracillima have central secondary ray but no central tertiary rays in branchlets, which are present in N. caespitosa . Because of the formation of central secondary ray, gametangia are not produced at first furcation of a branchlet. Additional differentiating characteristics are discussed below.

Nitella flagellifera differs from N. singaporensis by the absence of mucilage around gametangia and strongly reticulated oospore surface under LM ( Wood & Imahori 1965). According to other available descriptions of N. flagellifera from Brazil and China, the presence of abbreviated dactyls, less number of dactyls (2–3(–4)), rounded apex of penultimate cell in some populations, 2–4-furcate branchlets, 2–3 oogonia at a branchlet furcations (Chinese plants have unabbreviated dactyls, 2–3-furcate branchlets and solitary oogonia), occasional absence of central secondary ray substituted with antheridium, and absence of mucilage envelope of gametangia can be treated as differentiating traits with N. singaporensis ( Han & Li 1994; Bueno et al. 2016; Borges & Necchi 2018). The oospore surface in case of N. flagellifera is beaded imperfectly reticulated to reticulated in LM and in SEM ( Han & Li 1994; Borges & Necchi 2018), i.e. dissimilar with N. singaporensis .

Nitella gracillima differs by the 1–3-furcate branchlets, rare presence of gametangia at the first furcation of a branchlet, absence of mucilage around gametangia, and coarsely papillate to occasional imperfectly reticulated oospore surface ( Wood & Imahori 1965). According to the descriptions of other specimens from China and Japan, it has clearly confluent end cell of a branchlet, shorter first segment of the branchlet (less than a half of the latter), and its oospore surface is papillate in LM, papillate or beaded imperfectly reticulated in SEM ( Han & Li 1994; Sakayama et al. 2005).

Nitella dictyosperma differs by the absence of central secondary ray and mucilage around gametangia ( Wood & Imahori 1965). It has mainly 3-furcate branchlets, rare gametangia at first furcation of a branchlet and reticulated oospore surface. According to the description of specimens from China, N. dictyosperma has occasionally 4-furcated fertile branchlets, 3–4 tertiary segments of a branchlet, abbreviated dactyls, and reticulated oospore surface under LM ( Han & Li 1994), distinct from N. singaporensis .

It should be noted that sometimes descriptions and drawings referable to the same species from distant regions (e.g. Brazil and China) are significantly different from each other as well as with type specimen descriptions ( Wood & Imahori 1964, 1965), severely hampering critical comparison and raising the questions about species identity.

Figures

FIGURE 38. ML-phylogenetic tree inferred in PAUP with GTR+I+Г nucleotide substitution model from 89 rbcL sequences of Nitella. ML BP (> 50%) and BI PP (> 0.95) are shown. Thick branches received 100% BP and 1.00 PP support.The branch leading to the outgroup was 50% reduced in length. The newly obtained sequence of Nitella singaporensis is given in bold. Clade designation is according to Sakayama (2008).

FIGURES 5–19. The whorls and branchlets of Nitella singaporensis (all from isotype: LE): 5—apical part with mucilage recognised because of adhering of fine particles at its surface; 5–7—whorls of branchlets, showing long primary ray; 8–18—branchlets, showing variability in length of central secondary ray, from moderately elongated, i.e. exceeding the length of lateral secondary rays with their dactyls (8 & 9), nearly equal with them (10, 12, 14), to abbreviated, i.e. clearly shorter than lateral second segments (19), mainly spreading (8–16) or adpressed to the central one, resulting in last case in tassel-like general appearance of the branchlet (19); 8, 12, 15, 16—fine particles adhered to the surface of a branchlet, indicating presence of a mucilage at its surface; actually they are outlining the mucilage patches; 9—central secondary ray without furcation, appearing as long robust dactyl (arrowheads); 10 & 11—branchlets with central tertiary ray (arrowheads); 11—branchlet with central secondary ray with furcation (double arrowhead), producing central tertiary ray with 3 dactyls (arrowhead); 12—dactyls at the furcation at the top of central secondary ray (arrowhead); 13—central tertiary ray (arrowhead) at lateral secondary ray; 17—unfurcated lateral secondary rays (arrowheads), central secondary ray at the branchlet below is not clearly visible because of artificial deformation at its base. Scale: 5–11—2 mm, 12–17—1 mm, 18 & 19—0.5 mm.All photos by R. Romanov.

FIGURES 20–31. Nitella singaporensis from Singapore (isotype: LE), N. gracilis from Sicily (LE): 20 & 21—abbreviated central second segment (arrowhead), looking as an additional branchlet at first furcation; 21—rare case of 8 lateral secondary rays; 22—end parts of a branchlet with clearly visible central secondary ray (arrowhead); 23–25—2-celled dactyls, some of them appearing 1-celled after loss of end cell during storage in living state in the laboratory; 26—gametangia, [arrowhead—shortly stipitate antheridia, double arrowhead— production of antheridium in penultimate not end furcation], 27—oospore surface in LM, appearing as unclear reticulations, 28–30—stem nodes with easily visible nodal cell extruding from the base of branchlets below the whorl, resulting in nodal bulbil formation (30), 31—stem node of N. gracilis with nodal cell extruding from the base of branchlets and situated around it. Scale: 20–22—0.5 mm, 23–26, 28–31—0.2 mm, 27—20 μm. All photos by R. Romanov, except 26 by R.S.W. Yeo.

FIGURE 1–4. General appearance and whorl and branchlet of Nitella singaporensis: 1—pressed and dried material (holotype: SING); 2—pressed and dried material (isotype: LE); 3—reconstruction of general habit of branchlet whorl (view from above) from the photo of ethanol-fixed specimen (isotype: LE) showing mean of terms used for branchlet description: 1—primary ray, 2—lateral secondary ray, 3—central secondary ray, 4—lateral tertiary ray, 5—central tertiary ray, 6—dactyls, yellow—lateral secondary segments, green—central secondary segment, blue—central tertiary segment, orange—lateral tertiary segment; 4—fertile branchlet of rewetted herbarium specimen showing oogonia arrangement (isotype: LE). Scale: 1 mm. Scale: 2—1 cm, 4—1 mm. Photo 1 by D. Liew, drawing and other photos by R. Romanov.