Gynura mollis Schultz

Zhu, Li-Cui, Hariri, Muhammad Rifqi, Li, Cheng-Sheng & Ren, Chen, 2025, Clarification of the taxonomic and nomenclatural confusions with three Gynura names (Asteraceae: Senecioneae) based on Zollinger’s collections from Java, Indonesia, Phytotaxa 697 (3), pp. 219-233 : 228-232

Gynura mollis Schultz (in Zollinger 1845: 124, 127) ( Figs. 1–3 , 4A , 5 , 7–9 )

Type:— INDONESIA. East Java, Mt. Ardjuno, 1200–1800 m, September 1944, Zollinger H.2592 (lectotype designated here: P02562893 !; isolectotypes: BR0000029078073 !, G00604939 !, G00630048 !, G00630050 !, P02562892 !, and P02562894 !) ( Figs. 2 , 5 , 9 ) .

= Gynura aurantiaca var. robusta Zollinger & Moritzi (1845: 265) , syn. nov.

Type:— INDONESIA. East Java, Mt. Tengger , 2300 m, November 1944, Zollinger HZ.606 (lectotype here designated: P02562906 !; isolectotypes: BO1281963 !, P02562907 !, and P04389013 !) ( Figs. 4A , 7 , 8 ) .

= Gynura densiflora Miquel (1856: 99) , syn. nov.

Type:— INDONESIA. East Java, Mt. Tengger , 2300 m, November 1944, Zollinger HZ.606 [cited as “ H.2592 ” by Miquel (1856); see below] (holotype: P02562907 !; isotypes: BO1281963 !, P02562906 !, and P04389013 !) ( Figs. 4A , 7 , 8 ) .

= Gynura ajakensis Hochreutiner (1934: 330) , syn. nov.

Type:— INDONESIA. East Java, Mt. Tengger, Ajak-ajak, 2600 m, 20 January, 1905, Hochreutiner 2736 (holotype: G 00035628!) ( Fig. 1 ).

Notes:—We suggest that the gathering Zollinger H.2592 should include BR0000029078073 ( Fig. 9A ), G00604939 ( Fig. 2A ), G00630048 ( Fig. 2C ), G00630050 ( Fig. 2D ), P02562892 ( Fig. 5A ), P02562893 ( Fig. 5C ), and P02562894 ( Fig. 9C ), while Zollinger HZ.606 should contain BO1281963 ( Fig. 7A ), P02562906 ( Fig. 7C ), P02562907 ( Fig. 4A ), and P04389013 ( Fig. 8A ). In particular, P02562907 ( Fig. 4A ) needs some notes. It is the holotype of Gynura densiflora and cited as H.2592 by Miquel in the protologue. However, the specimen is actually annotated with both HZ.606 and H.2592 ( Fig. 4B ). Despite both gatherings undoubtedly representing the same species, P02562907, with coarser leaves more densely arranged along the stem and a more robust synflorescence, aligns more closely with the specimens of Zollinger HZ.606 (BO1281963, P04389013, and P02562906; Figs. 7 , 8 ). Consequently, we consider it better to assign P02562907 ( Fig. 4A ) to Zollinger HZ.606.

P 02562892 ( Fig. 5A ) and P 02562893 ( Fig. 5C ) are the two specimens Schultz relied on to describe G. mollis . Among them, P 02562893 ( Fig. 5C ) is better preserved and bears Schultz’s handwritten determination, and is selected as the lectotype of G. mollis .

Gynura aurantiaca var. robusta is also based on Zollinger HZ.606 ( Zollinger 1845). Among the four sheets of Zollinger HZ.606, P 02562906 ( Fig. 7C ) and BO 1281963 ( Fig. 7A ) were annotated on the determination slips as a potential variety of G. aurantiaca , although no varietal epithet was specified for P 02562906 ( Fig. 7D ) and a different epithet “ rugosa ” given for BO 1281963 ( Fig. 7B ). These should be the basis of var. robusta . P 02562906 ( Fig. 7C ) is here chosen as the lectotype because it is better preserved with an intact synflorescence.

Davies (1980) and Vanijajiva & Kadereit (2011) regarded G. densiflora as a superfluous name for G. aurantiaca . This is probably due to a misinterpretation of Miquel’s (1856) citation in the protologue of G. densiflora , where Miquel (1856) cited “ Gynura aurantiaca (haud DC.) Schultz Bip. in Herb. Zoll. ad n. 2592” following the morphological description. Davies (1980) may have assumed that Miquel cited legitimate G. aurantiaca as a synonym, thereby rendering G. densiflora superfluous. However, Miquel (1856) actually intended to indicate a misidentification of Zollinger H.2592. Therefore, Gynura densiflora is a legitimate name.

Backer (1940) once considered that Hochreutiner (1934), when describing G. ajakensis , mistakenly compared it to G. densiflora by erroneously adopting the original description of G. densifolia Miquel (1856: 99) as that of G. densiflora . Gynura densifolia is a completely different species, but simultaneously described with G. densiflora with a confusingly similar epithet, and it is listed immediately after G. densiflora . We, however, consider that this confusion is actually caused by two typographical errors in Hochreutiner’s (1934) text that mistook G. densifolia as G. densiflora (lines 25 and 30 on page 331). In the first paragraph of his discussion, Hochreutiner (1934: 331) clearly stated that he saw the type of G. densiflora (“Nous avons a Geneve, le type du G. densiflora ……”; Hochreutiner 1934: 331), and as a matter of fact, he annotated one of the sheets ( G 00604939; Fig. 2B ). In addition, he compared G. densiflora with G. ajakensis in the second paragraph. However, in the third paragraph, he paradoxically reclaimed not to have seen G. densiflora (“Nous n'avons pas vu le G. densiflora ……”; Hochreutiner 1934: 331) and re-compared G. densiflora with G. ajakensis . This inconsistency is difficult to reconcile without attributing it to typographical errors. Notably, Hochreutiner likely never examined the type of G. densifolia , which was absent from G, further supporting our hypothesis about the errors. Additional evidence comes from Hochreutiner’s (1934) reference to Koorders, who noted that the species was also found by Kuntze in Guedeh near Sindaglaja. This actually referred to the gathering Kuntze 4462. It was described by Kuntze (1891) as Crassocephalum densifolium var. atropurpureum Kuntze , but later synonymized by Koorders (1912) just with G. densifolia . Therefore, Hochreutiner (1934) actually intended to compare G. ajakensis with G. densifolia in the third paragraph of the discussion. Unfortunately, the typographical errors mistook G. densifolia as G. densiflora and led to the confusion.

Despite Backer’s (1940) misinterpretation of Hochreutiner’s (1934) discussion, we agree with Backer’s (1940) conclusion that G. ajakensis and G. densiflora are conspecific. Hochreutiner’s (1934) considered G. ajakensis different from G. densiflora in having denser synflorescences and different types of leaves and indument. However. we do not observe significant differences in leaf shape or indument. While the synflorescence on the type of G. ajakensis is indeed denser compared to those on the specimens of Zollinger H.2592 kept at G ( Fig. 2 ), which Hochreutiner (1934) regarded as the “ type ” of G. densifolia , these variations fall well within the normal range of this species. The capitula are generally densely arranged, especially in comparison to G. aurantiaca , but they also exhibit some variability, with peduncles elongating to a certain extent.

Gynura mollis is clearly most similar to G. aurantiaca . Their differences have been well summarized by Backer (1940). Gynura aurantiaca is a climbing or creeping plant characterized by purple hairs on young parts. The leaves are largely lyrate and often have conspicuous auricles at the base, and the capitula are generally loosely arranged. In contrast, G. mollis is an erect plant with white hairs on young parts, undivided leaves lacking auricles at the base, and generally dense synflorescences. In addition, G. mollis is restricted to East Java, whereas G. aurantiaca is a common species in Sumatra, West and Central Java, only slightly overlapping the distribution range of G. mollis in the westernmost part of East Java (Mts. Lawu and Wilis).

Figures

FIGURE 1. Holotype of Gynura ajakensis, Hochreutiner 2736 (G00035628, with two sheets; A, B) from Mt. Tengger in East Java, Indonesia.

FIGURE 2. Three specimens of Zollinger H.2592 kept at G.A. G00604939.B.Annotations made by Hochreutiner (lower slip) and Davies (upper slip) on G00604939. C. G00630048 (two sheets at G, but only one shown here). D. G00630050.

FIGURE 3. A potential original material of Gynura densiflora kept at U. A. Specimen (U0001285). B. Annotations on U0001285. Note that the annotation of “prob. Z.M. 2592” (green arrow) was probably made by Backer as indicated in his paper (Backer 1940).

FIGURE 4. Two specimens of Zollinger H.2592 from the Drake Herbarium. A. P02562907; the holotype of Gynura densiflora. B. Annotations on P02562907, showing the indicated collection locality “Bandong” (green arrow), Miquel’s annotations (“Gynura aurantiaca (non DC.) Schultz” and “Gynura densiflora Miq.”; middle slip), and the additional collection number “606Z” (= HZ. 606; bottom slip). C. P02562905. D. Annotations on P02562905, showing the indicated collection locality “Mt. Ardjuno” (green arrow).

FIGURE 5. Syntypes of Gynura mollis from the Schultz Herbarium. A. P02562892. B. Annotations on P02562892, showing two distant localities: “Mt. Tankuban” and “Mt. Ardjuno” (green arrows). C. P02562893. D. Annotations on P02562893, showing Schultz’s determination. Note that the two specimens are numbered as both “H.2592” and “HZ.442”, but Schultz only cited “H.2592” in the protologue of G. mollis. He also specifically annotated the single individual on P02562892 as “2592” (B, red arrow) and made a “wrong” mark (D, blue arrow) on the number of “HZ.442” on P02562893, suggesting that Schultz noticed the mixture between Zollinger H.2592 and HZ.442.

FIGURE 7. Two specimens of Zollinger HZ.606. A. BO1281963. B. Annotations on BO1281963. C. P02562906. D. Annotations on P02562906. Both specimens are indicated with the collection locality “Mt. Tengger” (green arrows) and suggested as a variety of Gynura aurantiaca (red arrows).

FIGURE 8. An additional specimen of Zollinger HZ.606. A. P04389013. B. Annotations on P04389013.

FIGURE 9. Two specimens of Zollinger HZ.442. A. BR0000029078073. B. Annotations on BR0000029078073, showing the indicated locality “Bandong” (green arrow). C. P02562894. D. Annotations on P02562894, showing the indicated locality “Mt. Tankuban” (green arrow).