Ann. Bot. Fennici 37: 249–253 ISSN 0003-3847
Helsinki 15 December 2000 © Finnish Zoological and Botanical Publishing Board 2000
A new species of Doronicum (Asteraceae,
Senecioneae) from central China
Inés Álvarez Fernández & Gonzalo Nieto Feliner
Álvarez Fernández, I. & Nieto Feliner, G., Real Jardín Botánico, CSIC, Plaza de Murillo
2, E-28014 Madrid, Spain
Received 2 May 2000, accepted 31 July 2000
Álvarez Fernández, I. & Nieto Feliner, G. 2000: A new species of Doronicum
(Asteraceae, Senecioneae) from central China. — Ann. Bot. Fennici 37: 249–253.
The name Doronicum thibetanum Cavill. has been inadvertently applied in herbarium
labels to designate a species distributed in central China (provinces of Gansu and
Sichuan). However, the study of the type specimen reveals that it is actually a species of
Aster. The central Asian species of Doronicum has no available name and is thus here
described as D. cavillieri Álv. Fern. & Nieto Fel. This new species has as its closest
relatives ﬁve central Asian species: D. altaicum Pall., D. briquetii Cavill., D. falconeri
C.B. Clarke, D. kamaonense (DC.) Álv. Fern. and D. stenoglossum Maxim. Their mor-
phological afﬁnities and phylogenetic relationships are commented upon and an
identiﬁcation key is provided.
Key words: Asteraceae, China, Doronicum, Senecioneae, new species, taxonomy
The genus Doronicum (Asteraceae, Senecioneae)
is comprised of 26 species and four subspecies
distributed in Asia, Europe and north Africa (I.
Álvarez Fernández unpubl.). All representatives
are perennial rhizomatose herbs with alternate
simple leaves, bearing one to several radiate yel-
low-ﬂowered capitula. The involucre is composed
of 2–3 rows of uniform herbaceus phyllaries. The
cypselae are broadly elliptic-oblong and have 10
longitudinal ribs. The pappus consists of bristles
in all cypselae except in those of the radiate ﬂowers
of the heterocarpic capitula, where it is lacking.
Doronicum species occur in open or forested habi-
tats from sea level up to 5 000 m of elevation.
The genus has been traditionally placed in the
tribe Senecioneae (Cassini 1819, Bentham 1873,
Hoffmann 1892, Nordenstam 1977, Bremer
1994). This placement has received molecular
support from the chloroplast gene ndhF (I. Álvarez
Fernández, Fuertes Aguilar & Nieto Feliner
unpubl.). Nevertheless, the circumscription of the
genus has undergone signiﬁcant changes. The
most signiﬁcant has been the exclusion of species
currently placed in several other genera (Arnica,
250 Álvarez Fernández & Nieto Feliner •ANN. BOT. FENNICI 37
Aster, Nannoglottis), belonging to different tribes.
A taxonomic revision of the genus based on
ca. 4 500 herbarium specimens from 47 herbaria
has resulted in a new taxonomic treatment
(Álvarez Fernández & Nieto Feliner 1999 and I.
Álvarez Fernández unpubl.). The objective of the
present paper is to provide a formal taxonomic
description for an undescribed Chinese species.
This species had been identiﬁed in herbarium la-
bels as Doronicum thibetanum Cavill. However,
to our knowledge the application of Cavillier’s
name to populations of the species we are here
describing has never been published. Further, such
application is wrong because the study of the type
material of D. thibetanum (Cavillier 1907) reveals
that the type specimen is actually an Aster. Addi-
tionally, a discussion is provided on the relation-
ships among the central Asian species of the ge-
nus where the closest relatives of the new species
are found. This is based not only on the taxonomic
revision but also on a phylogenetic study using
DNA sequences and morphological data (I.
Álvarez Fernández, J. Fuertes Aguilar & G. Nieto
MATERIAL AND METHODS
More than 50 morphological qualitative and quan-
titative characters were studied from the 19 dried
specimens of the new species found in the her-
baria examined. These specimens are from BM,
E, K, GH, LE, NY, MO, S, and W. Observations
were made using either a binocular lens or SEM.
Quantitative characters were recorded with the aid
of a Brown & Sharpe Plus digital caliper (model
RESULTS AND DISCUSSION
DESCRIPTION OF THE NEW SPECIES
Doronicum cavillieri Álv. Fern. & Nieto Fel.,
sp. nova (Fig. 1)
Speciei D. briquetii Cavill. similis, a qua prae-
sertim differt bracteis involucralibus apice glan-
dula sessili praeditis, foliis caulinis constanter
integris atque obtusis, capitulo basi pilis glan-
duliferis ut plurimum 2 mm longis instructo (nec
ut in D. briquetii ut plurimum 5 mm longis).
HOLOTYPE: China. SW Gansu, T’ao river basin, Minshan
range, mt. Kuang ke, 3 700 m, meadows, VI.1925, J. F.
Rock s. n. (NY; isotypes BM, GH, LE).
Rhizome ﬂeshy to somewhat woody, glabrous,
with uniform internodes. Stem up to 30 cm, not
branched, sometimes with persistent leaf remains
forming dark scales at base, with uniform
internodes shorter than the adjacent leaves or
scarcely longer than them. Leaves distributed
along the stem, entire, with blunt apex, truncate
to attenuate base, and pinnate-reticulated veins.
Basal leaves orbicular to elliptic, blade 14–35 ×
15–30 mm, petiole 30–87 × 0.75–1.5 mm, some-
times reduced (cataphylls). Cauline leaves sessile,
semiamplexicaul, broadly elliptic to ovate, 15–
50 × 7–35 mm. Capitula terminal, solitary, 30–60
mm in diameter including ligules. Involucre 20–
35 mm in diameter. Phyllaries subulate to nar-
rowly ovate, 10–16 × 1.5–2.5 mm, with blunt apex
that bears a sessile gland. Receptacle glabrous.
Flowers yellow. Style 4.0–6.25 mm. Stamen ﬁla-
ments ca. 4 mm, anthers ca. 2.5 mm. Ligulate
ﬂowers 20–25 × 2–3 mm, very narrowly elliptic,
the ligules with 2–3 teeth at apex. Tubular disk
ﬂowers ca. 5 × 2 mm. Cypselae homomorphic,
ca. 3 × 1 mm (immature), brown. Pappus ca. 4
mm, with minutely scabrous bristles, whitish to
yellowish. Indumentum variable, composed of
glandular hairs up to 2 mm and uniseriate
eglandular hairs up to 0.5 mm, abundant near the
capitulum to very scarce at the base of the plant,
eglandular hairs present mainly on the leaves.
Flowering in June. Mature cypselae not seen.
DISTRIBUTION: Mountains of central China
(provinces of Gansu and Sichuan).
HABITAT: Open rocky places, grassy slopes,
meadows, and in forests, from 3 000 m to 3 700 m
ADDITIONAL SPECIMENS EXAMINED (PARATYPES): — China.
Gansu Province. Siku, high cool gullies, 16.VI.1914 R.
Farrer 144 (E); T’ao river basin, Merku valley, SW Gansu,
3 000 m, grassy slopes, VI.1925 J. F. Rock 12192 (E, K,
NY, S, W); T’ao river basin, Minshan range, mt. Kuang
Ke, SW Gansu, 3 700 m, meadows, VI.1925 J. F. Rock
12389 (BM, GH, LE, NY); upper Tebbu country, foot of
Shimen, SW Gansu, 3 660 m, rock limestone wall extend-
ing east & west, 1925 J. F. Rock 13020 (E, K, GH, LE, NY,
ANN. BOT. FENNICI 37 •A new Doronicum from China 251
S, W). Sichuan Province. Scarp on S edge of Hongyuan
plain, on road to Songpan, 3 350 m, in moss under shrubs,
1989 D. Chamberlain, P. Cox & P. Hutchison (E); Lian-
huashan, Kangle Xian, 3 000 m, under forest, 23.VI.1991
G. H. Wang 91161 (MO).
CENTRAL ASIAN RELATED SPECIES
According to our taxonomy (I. Álvarez Fernández
unpubl.), there are six species of Doronicum oc-
curring east of the 60° meridian: D. altaicum Pall.,
D. briquetii Cavill., D. falconeri C.B. Clarke, D.
kamaonense (DC.) Álv. Fern. (= D. roylei DC.),
D. stenoglossum Maxim., and the new species
described here as D. cavillieri. Their distribution
areas span a large territory from Altai to the east-
ern Himalayas and central China across Pamir and
Kashmir. Some of their areas partly overlap.
Although we have failed to detect a synapo-
morphic morphological character that supports
Álv. Fern. & Nieto Fel.
— a and b: Habit. — c: Ca-
pitulum. — d: Involucral
bract. — e: Apex of an in-
volucral bract. — f: Ligul-
ate ﬂower. — g: Tubular
ﬂower. — h: Androecium.
252 Álvarez Fernández & Nieto Feliner •ANN. BOT. FENNICI 37
their monophyly, they all posses a combination
of non-exclusive characters. These are the gla-
brous and somewhat woody rhizomes, the uni-
form distribution of leaves along the stem, and
the pinnate-reticulate leaf venation. A parsimony
analysis of the genus, based on three different data
sets (nuclear ribosomal ITS sequences, chloroplast
spacer sequences of the trnL-trnF region, and
morphological characters) gives some support to
the monophyly of the group composed of these
central and eastern Asian species (I. Álvarez
Fernández, J. Fuertes Aguilar & G. Nieto Feliner
unpubl.). In the phylogenetic hypothesis obtained
by simultaneously analyzing the three data sets
for representatives of all species of Doronicum,
the six Asian species form a clade. However, the
bootstrap support for this clade is moderate. Fur-
ther molecular evidence is needed to determine if
these species do form a monophyletic group and
thus deserve taxonomic recognition at an infra-
generic rank. Within that group, D. cavillieri ap-
pears as sister to D. briquetii, and this pair, in turn,
is sister to the one formed by D. falconeri, D.
kamaonense and D. stenoglossum.
At the speciﬁc level, taxonomic differentia-
tion is not problematic. Characters relative to the
indumentum, homocarpy, and number of capitula
are useful to distinguish Doronicum altaicum, D.
briquetii, D. falconeri and D. kamaonense. The
remaining two species, D. stenoglossum and D.
cavillieri, additionally present some exclusive
On morphological basis, the closest species
to Doronicum cavillieri are D. altaicum and D.
briquetii. The three of them are homocarpic and
bear a single capitulum (D. altaicum exception-
ally up to four). Besides the exclusive character
of D. cavillieri (sessil gland in the apex of every
involucral bract), this species can also be distin-
guished from the other two by the type of
indumentum. Glandular hairs at the base of the
capitula are at most 2 mm long in D. cavillieri
while those in D. briquetii are up to 5 mm, and
the ones in D. altaicum are almost sessile (< 1
Of the remaining Asian species, Doronicum
stenoglossum can be easily distinguished from D.
cavillieri. In fact, D. stenoglossum was placed by
Cavillier (1911) in a different section (sect.
Soulieastrum) than the bulk of the genus (sect.
Doronicastrum). Characters separating the two
species are the number of heads (usually more
than two in D. stenoglossum), the shape of ligules
(almost linear, 0.5–1.5 mm wide in D. steno-
glossum vs. narrowly elliptic, 2–3 mm wide in D.
cavillieri), and the relative length of ligules as
compared to involucral bracts (clearly longer in
D. cavillieri vs. almost equal or even shorter in
D. stenoglossum). Doronicum kamaonense dif-
fers from D. cavillieri mainly by its heterocarpy
and a higher number of capitula (2–18).
KEY TO THE SPECIES OF DORONICUM IN CENTRAL
1. Heterocarpic heads; cypselae of the ligulate ﬂowers
without a pappus ...................................................... 2
1. Homocarpic heads; pappus present in all cypselae
although sometimes caducous ................................. 3
2. Unbranched stem with a solitary terminal capitulum
(40–75 mm in diameter, ligules included) D. falconeri
2. Branched stem with several (2–18) capitula (15-40 mm
in diameter, ligules included) ........... D. kamaonense
3. Flowers pale yellow to greenish; ligules linear (0.5–
1.5 mm wide); involucral bracts equaling or even
exceeding ligules; pappus of ligulate ﬂowers sometimes
caducous ......................................... D. stenoglossum
3. Flowers yellow; ligules narrowly elliptic (1.5–3.5 mm
wide); involucral bracts much shorter than ligules, rarely
equaling them; pappus never caducous ................... 4
4. Involucral bracts with a blunt apex (ending in a sessile
gland) ..................................................... D. cavillieri
4. Involucral bracts with an acute apex (without a gland
at the apex)............................................................... 5
5. Indumentum at the base of the head usually composed
of glandular hairs (1–5 mm); capitulum solitary ........
................................................................ D. briquetii
5. Indumentum at the base of the head composed of
subsessile glands or hairs (less than 1 mm), sometimes
subglabrous; capitula 1–4 ....................... D. altaicum
NOMENCLATURE OF DORONICUM CAVILLIERI
Doronicum thibetanum was described by Cavillier
(1907). The author cited a single collection in the
protologue: “Thibet (Murr, ann. 1882, in herb.
Boissier)”. In G-BOIS there is only one sheet
whose label matches the protologue: “Tibet, Murr
ANN. BOT. FENNICI 37 •A new Doronicum from China 253
1882”. Attached to this specimen is another label
showing Cavillier’s handwriting: “Doronicum
thibetanum sp. nov., Fr. Cavillier determ, anno
1906”. There is thus a single type element on
which the identity of the name has to be based.
The overall appearance of the specimen is rather
ambiguous and the ligules are currently yellow-
ish. But this latter character is rather unstable and
thus the original ligule color may well be blue
(Grierson 1964). An examination of the style
branches is conclusive to reject the identiﬁcation
of the type specimen as a Doronicum.
The ﬂattened acuminate style branches com-
bined with the lack of anther appendages and the
radiate capitula indicate that the type of Doron-
icum thibetanum belongs in tribe Astereae. Char-
acters such as a small herb with one capitulum,
herbaceous subequal (> 1 mm wide) phyllaries,
relatively wide ligules (> 1 mm wide), and a sim-
ple pappus as long as the corolla suggest an Aster
sect. Alpigeni subsect. Homochaeta (Grierson
1964). A more precise identiﬁcation is neverthe-
less problematic due to the immature state of the
specimen as well as to the lack of other relevant
characters of the rhizome.
Two factors contributed to all this taxonomic
confusion. First, the fact that Cavillier misidentiﬁed
an Aster for a Doronicum. This was probably due
to the misleading color of the ligules and the fact
that he did not dissect the capitulum (the only one
in the specimen was intact). Second, the existence
of a good species of Doronicum in the Tibet area
demanding a name coupled with an available epi-
thet associated with an ambiguous description.
ACKNOWLEDGEMENTS: We are grateful to the curators of the
herbaria BM, E, K, G-BOIS, GH, LE, NY, MO, S, and W
for the loan specimens, Ian Hedge for providing informa-
tion, Manuel Laínz for the latin diagnosis, Juan Castillo for
the drawings, and Bernardo Fernández for technical assist-
ance. This work has been supported by grant DGES
PB96-0849 of the Spanish Dirección General de Enseñanza
Superior e Investigación Cientíﬁca.
Álvarez Fernández, I. & Nieto Feliner, G. 1999: Lecto-
typiﬁcation of 16 species names in Doronicum (Astera-
ceae, Senecioneae). — Taxon 48: 801–806.
Bentham, G. 1873: Notes on the classiﬁcation, history, and
geographical distribution of Compositae. — J. Linn.
Soc., Bot. 13: 335–577.
Bremer, K. 1994: Asteraceae. Cladistics & Classiﬁcation.
— Timber Press, Portland, Oregon. 752 pp.
Cassini, H. 1819: Suite du sixième mémoire sur la famille
des synanthérées. — J. Phys. Chim. Hist. Nat. Arts. 88:
Cavillier, F. 1907: Étude sur les Doronicum a fruits homo-
morphes. — Annuaire Conserv. Jardin Bot. Genève 10:
Cavillier, F. 1911: Nouvelles études sur le genre Doronicum.
— Annuaire Conserv. Jardin Bot. Genève 13–14: 195–
Grierson, A. J. C. 1964: A revision of the Asters of the
Himalayan area. — Notes Royal Bot. Garden Edinburgh
Hoffmann, O. 1892: Doronicum L. — In: Engler, A. &
Prantl, K (eds.), Die Natürlichen Pﬂanzenfamilien 4:
Nordenstam, B. 1977: Senecioneae and Liabeae systematic
review. — In: Heywood, J. B. & Turner, B. L. (eds.),
The biology and chemistry of the Compositae: 799–
830. Acad. Press, London.