Andryala subglabrata (DC) M.Z.Ferreira, Alv.Fern. & M.Seq.

Ferreira, Maria Zita, Fernández, Inés Álvarez & Sequeira, Miguel Menezes De, 2023, Taxonomic and nomenclatural notes on Andryala glandulosa subsp. cheiranthifolia (Cichorieae, Asteraceae) from Madeira Island, Portugal, Phytotaxa 579 (4), pp. 243-254 : 245-248

publication ID	https://doi.org/ 10.11646/phytotaxa.579.4.2
DOI	https://doi.org/10.5281/zenodo.7568632
persistent identifier	https://treatment.plazi.org/id/03BA8783-FFA4-FFF3-FF7F-FC56FF18B477
treatment provided by	Plazi
scientific name	Andryala subglabrata (DC) M.Z.Ferreira, Alv.Fern. & M.Seq.
status

Treatment

Andryala subglabrata (DC) M.Z.Ferreira, Alv.Fern. & M.Seq. View in CoL , stat. & comb. nov.

≡ Andryala varia var. subglabrata View in CoL Candolle (1838: 246)

Type:— PORTUGAL. Madeira. In rupibus subumbrosis convallium Maderae ad alt. 2000–5000 ped. ubique vulg., 1832, M. Low. 399 (lectotype G00326360 ex herb. DC., designated here) .

Andryala varia var. aprica Candolle (1838: 246) .

Type:— PORTUGAL. Madeira. Hab. in Maderae ad altit. 3000–5000 ped. in locis rupibus apricis siccis, magis quam quibus subv. 1 nascitur, 1832, M. Low. 388 (holotype G00326377 ex herb. DC., specimen on the left) .

Andryala cheiranthifolia subv. integrifolia Lowe (1868: 563) View in CoL .

Type:— PORTUGAL. Madeira. Rocks in Ribeiro Frio , 29 June 1827, Lowe 130 (lectotype K000251909!, designated here) .

Andryala cheiranthifolia subv. runcinata Lowe (1868: 564) View in CoL .

Type:— PORTUGAL. Madeira. Rock at the head of the Rib. Funda above the Lagoa or Fanal , 18 July 1855, R. T. Lowe 400 XX (lectotype BM000072531a!, designated here; isolectotype: K000251919!) .

Description:—Annual to biennial herbs, sometimes perennating, multi- or single-stemmed ( Fig. 1a View FIGURE 1 ). STEMS 37– 94(–105) cm, branched in the upper half, less frequently from the base or in the upper third, puberulous to tomentose with stellate hairs and numerous glandular hairs only in the upper part. LEAVES glaucous to deep green, puberulous to tomentose on both faces with stellate hairs mixed with ± abundant glandular hairs on the upper leaves; lower leaves arranged in a rosette, often marcescent at anthesis, 83–223 × 9–47(–68) mm, attenuate into a winged petiole 16–50 mm, usually lanceolate, apex acute, and margin subentire to pinnatipartite with lobes more than 3 mm wide ( Fig. 1b View FIGURE 1 ); cauline leaves 55–130(–152) × 5–41 mm, semiamplexicaul, lanceolate to linear-lanceolate, base ± rounded sometimes attenuate, apex acute sometimes acuminate, and margin subentire to pinnatipartite; upper leaves 14–44 × 2–10 mm, semiamplexicaul or amplexicaul, ovate-lanceolate to narrowly ovate-lanceolate, base ± rounded sometimes cuneate, apex acuminate, and margin entire, rarely subentire to dentate. INFLORESCENCE corymbose or paniculate-corymbose, with 5–12 capitula. CAPITULA (12–) 16–30 mm in diameter ( Fig. 1c View FIGURE 1 ); peduncles 23–65(–77) mm with stellate hairs and some glandular hairs 0.9–1.8 mm; involucre 8–12 × 9–18 mm, campanulate at anthesis, with involucral bracts in 2–3 rows; external involucral bracts 5–9 × 0.7–1.5 mm, lanceolate to linear-lanceolate, apex frequently acuminate, involute enfolding a floret, the outer face pubescent-tomentose with stellate hairs and glandular hairs 0.8–1.8 mm, yellowish at the apex and blackish at the base, the inner face glabrous; internal involucral bracts 4–7 × 1–2 mm, with broad scariose margins or almost completely scariose, receptacle convex, villous with long setose hairs 3–6 mm (2 to 4 times longer than the cypselae). FLORETS ligulate, golden yellow, the external with a tube of 2.5–5 mm and ligule of 6–12 × 1.5–4.5 mm ( Fig. 1d View FIGURE 1 ). CYPSELAE 1–1.7 × 0.4–0.6 mm ( Fig. 1e View FIGURE 1 ), usually obconical dark brown with white ribs, apex with an inner ring of teeth equalling or slightly exceeding the prolongation of the ribs; pappus of white bristles 4–6 mm, denticulate at the base.

Ecology:— Andryala subglabrata occurs between 500–1820 m, growing mainly on dry rocky soils. It also grows in shady and humid places of the stink-laurel forest, on rocky substrates (on the north face of Madeira). It can also be found alongside roads.

Conservation status: —This species is common in Madeira and should be listed as Least Concern (LC) according to the IUCN Red List categories and criteria ( IUCN 2012).

Distribution: —Although Andryala subglabrata is found throughout the island of Madeira, it occurs mostly in inland areas such as Serra d’Água, S„o Vicente (especially in the vicinities of Encumeada), Paul da Serra, Rabaçal, Curral das Feiras, Camacha, and Pico do Areeiro ( Fig. 2 View FIGURE 2 ). However, it also grows in localities near the southwest coast of Madeira at high elevations (namely Prazeres and Faj„ da Ovelha). Unlike A. glandulosa , it does not occur in Porto Santo Island or in the Desertas Islands.

Taxonomic remarks: —L´Héritier (1785) included an illustration of A. cheiranthifolia in the protologue. However, a specimen from Candolle´s collection, currently deposited at the Geneva herbarium (G00470810) under Andryala cheiranthifolia , was also traced ( Fig. 3A View FIGURE 3 ). According to the label, it belonged to L’Heritier’s herbarium (“ ach. de l’herb. de L’Héritier ”) and corresponds to the original material (“ type de l’espèce ”).An additional label, handwritten by Candolle, reads “ A. cheiranthifolia L’Hér. ”, accompanied by the synonyms “ Andryala glandulosa Lam. ” and “ Andryala varia ε cheiranthifolia DC. ” Candolle (1838) clearly stated that his description of A. varia var. ε cheiranthifolia was based on a specimen from L´Heritier’s herbarium. In effect, Candolle acquired L’Heritier´s herbarium and made it the basis of his own collection, now deposited at G ( Stafleu & Cowan 1981). Another type specimen was located at the British Natural History Museum, under the barcode number BM001124974, accompanied by a label indicating its origin (“ Hort. Paris L’Héritier ”). The label was handwritten by Joseph Banks, an English botanist who collected many plants in Madeira during Captain Cook´s first circumnavigation voyage ( Hansen 1980) and maintained correspondence with L’Héritier ( Dawson 1958). An additional annotation in pencil reads “ Andryala cheiranthifolia L. Her. ! ” Thus, it is possible that this specimen originally belonged to L’Heritier’s collection, considering still that L’Héritier remained for 15 months in London between 1786 and 1787, a trip which resulted in the description of 19 new Macaronesian endemics ( Francisco-Ortega et al. 2008). Although both specimens match the original description, the most complete specimen, G00470810, is here selected as the lectotype of Andryala cheiranthifolia L’Hér. Morphologically these two specimens resemble A. glandulosa (i.e. A. glandulosa subsp. glandulos a) in several characters (e.g. stem and leaf glandular indumentum, leaf margin, insertion and shape, peduncle diameter, involucre shape at anthesis). Moreover, both resemble morphologically the lectotype of A. glandulosa ( Fig. 3C View FIGURE 3 ) designated by Ferreira et al. (2015b).

In conformity with the protologues, both A. glandulosa and A. cheiranthifolia exhibit chaffy-like inner involucral bracts, which are distinct from the outer involucral bracts. Lamarck (1783: 154) highlighted this feature when describing A. glandulosa : “… leur réceptacle est légérement velu dans son centre, & muni de quelques paillettes, à sa circonférence. Ces paillettes, quoique voisines du calice, sont tout-à-fait distinguées de ses folioles ”. Likewise, L’Héritier (1785: 36) underlined the presence of straw-like bracts surrounding the receptacle in A. cheiranthifolia (“ Receptaculum planum, villosum, cinctum paleis … ”), stressing also its numerous involucral bracts (“ CALYX … squammis plurimis … ”). Indeed, A. glandulosa (presently known as A. glandulosa subsp. glandulosa ) displays 4(– 5) rows of involucral bracts, the inner ones chaffy and glabrous and the outer ones densely stellate and glandular tomentose, each enfolding completely a cypsela.

In view of the abovementioned, A. cheiranthifolia should be treated as a synonym of A. glandulosa , and the latter is to be considered as the correct name. Candolle (1838), although with an expression of doubt and therefore without nomenclatural relevance, already included, besides A. cheiranthifolia , also A. glandulosa in his A. varia var. cheiranthifolia .

Fernandes (1959) stated that her study of herbarium specimens supports the recognition of two distinct species in Madeira: one from the coastal areas, A. glandulosa (= A. robusta ), and another from the mountainous regions, A. cheiranthifolia L´Hér. (= A. varia Lowe ex DC. p.p., according to the author). However, Fernandes (op. cit.), in agreement with Lowe, preferred to classify these two taxonomic entities as subspecies of A. glandulosa . Indeed, Lowe (1868) recognized a predominantly maritime form of the species, A. cheiranthifolia var. α congesta, and a mainly mountain form, A. cheiranthifolia var. β sparsiflora. Our study of herbarium material and specimens collected during our fieldwork showed that the plants included in A. glandulosa subsp. cheiranthifolia fit the original description of A. cheiranthifolia var. β sparsiflora. Moreover, Lowe (1868: 562, 563) distinguished the two varieties based on the involucral bracts: “ scales of inv. fewer or more remote and distinct … ” ( A. cheiranthifolia var. β sparsiflora) and “ outer scales of inv. linear lanceolate herbaceous thickly tomentose and fulvo-glandulose, the inner chaffy naked ” ( A. cheiranthifolia var. α congesta). The study of the original material of A. cheiranthifolia var. β sparsiflora (“ A. pinnatifida and A. denudata Sol. ! in BH ”), located at the BM herbarium under the barcode numbers BM001050371 and BM000829754, showed that both have stems with few stellate hairs and abundant glandular hairs only in the upper part, rather lax inflorescences, slender peduncles, campanulate involucres at anthesis, and few involucral bracts. These features are clearly distinct from those exhibited by L’Heritier’s A. cheiranthifolia . Additionally, Lowe (op. cit.) synonymized A. cheiranthifolia with A. cheiranthifolia var. β sparsiflora subvar. runcinata Lowe. However , our examination of the type material (BM000829755) showed that A. cheiranthifolia var. β sparsiflora subvar. runcinata does not correspond to A. cheiranthifolia . Consequently, the plants treated by Fernandes as A. glandulosa subsp. varia and by later authors under A. glandulosa subsp. cheiranthifolia do not represent A. cheiranthifolia in the sense of its type and, moreover, should be recognized at species rank. Considering that Andryala varia var. subglabrata DC. is the earliest legitimate name for this taxon, we propose a new combination: A. subglabrata . A specimen from the Candolle collection, G00326360, is here selected as lectotype. It was gathered in 1832 by Lowe and is labelled “ A. varia var. α subvar. 1 nob.”. Morphologically it resembles a specimen from Lowe´s collection under A. cheiranthifolia var. β sparsiflora subv. runcinata (BM000072531a).

Scopoli described A. tomentosa Scopoli (1787: 12–13) based on cultivated plants of uncertain origin. Several authors (e.g. Aiton 1789, Candolle 1838) placed it in A. cheiranthifolia , along with A. glandulosa . In the protologue, Scopoli (1787) stressed that the lower leaf margin is profoundly pinnatifid (“ Folia diversa ; nunc enim quae ex radice oriuntur profunde pinnatifida , segmentis lanceolatis, patulis, flexuosis, obtusis; nunc vero omnia sunt lanceolata; imis tamen sinuato-dentatis: superioribus autem integris.”), a feature also observed in A. subglabrata (contrary to A. glandulosa with lower leaves entire to subentire or sometimes lobate-pinnatifid). The description of A. tomentosa also fits A. subglabrata in what concerns the number of rows of the involucral bracts (“ Calyx communis subrotundus , constans unica serie foliolorum linearium …”). Hence, it is possible that A. tomentosa and A. subglabrata correspond to the same taxonomic entity. No original material of A. tomentosa was located to clarify this issue. In fact, Scopoli’s herbarium and types were mainly lost or destroyed during the World War II ( Stafleu & Cowan 1985). Moreover, A. tomentosa Scop. is a later homonym and, therefore, illegitimate [ A. tomentosa Chaix in Villars (1786: 366) was validly published a year before]. Considering the above, the adoption of A. subglabrata instead of A. tomentosa is acceptable.

Andryala subglabrata is morphologically distinct from A. glandulosa ( Fig. 4A, 4B View FIGURE 4 ). It differs mainly by the habit, stem and leaf indumentum, lower and cauline leaf margins, glandular indumentum of the peduncle, inflorescence, capitula diameter, involucre size and shape, as well as number of rows of the involucral bracts and indumentum ( Tab. 1 View TABLE 1 ).