Available via license: CC BY 4.0
Content may be subject to copyright.
Taxonomy of an endemic Aristolochia (Aristolochiaceae)
from the Iberian Peninsula
by
Andrea Costa
Departamento de Biodiversidad, Real Jardín Botánico, CSIC, Plaza de Murillo 2, 28014 Madrid, Spain. andrea.baccello@gmail.com
Abstract
Costa, A. 2008. Taxonomy of an endemic Aristolochia (Aristolo-
chiaceae) from the Iberian Peninsula. Anales Jard. Bot. Madrid
65(2): 173-178.
The taxonomy of an Iberian endemic Aristolochia is treated,
based on morphological and cytological characters. A brief de-
scription of its main diagnostic characters, distribution and habi-
tat is included, as well as a distribution map and a few com-
ments on its possible phylogenetic relationships. A new combi-
nation is proposed, raising this taxon from subspecies to a
species proper: A. castellana (Nardi) Costa. A revised dichoto-
mous key for the Iberian taxa of the genus is proposed.
Keywords: Aristolochia pallida subsp. castellana, Aristolochia
castellana, morphology, acid plutonic rock, granite, dichoto-
mous key for the Iberian Aristolochia species.
Resumen
Costa, A. 2008. Taxonomía de una Aristolochia (Aristolochia-
ceae) endémica de la Península ibérica. Anales Jard. Bot. Madrid
65(2): 173-178 (en inglés).
El presente trabajo trata la taxonomía de una Aristolochia endé-
mica de la Península Ibérica, basándose en caracteres morfológi-
cos y citológicos. Se incluye una breve descripción de los carac-
teres diagnósticos principales, de su distribución y hábitat, así
como un mapa de su distribución y algunos comentarios sobre
sus posibles relaciones filogenéticas. Se propone una nueva
combinación, ascendiendo el taxon de subespecie a especie: A.
castellana (Nardi) Costa. Se propone una nueva clave dicotómi-
ca para los táxones del género presentes en la Península Ibérica.
Palabras clave: Aristolochia pallida subsp. castellana, Aristolochia
castellana, morfología, roca plutónica ácida, granito, clave dicotó-
mica para las especies ibéricas de Aristolochia.
Anales del Jardín Botánico de Madrid
Vol. 65(2): 173-178
julio-diciembre 2008
ISSN: 0211-1322
Introduction
During a trip in central Spain, in 1986, an interest-
ing new taxon of Aristolochia from the Iberian Penin-
sula was discovered (Nardi, 1988; Ball & al., 1993).
The overall vegetative appearance, with the exception
of a general reduced size of all parts, strikingly resem-
bled A. lutea Desf., a common Central-Eastern Mediter-
ranean species (Ball & al., 1993). A. lutea belongs to the
group A. pallida, together with A. pallida Willd., a main-
ly Central Mediterranean species, and a few endemic
species from the Central-Eastern Mediterranean.
The chromosome number turned out to be 2n = 10,
different from A. lutea (2n = 8; Fabbri & Fagioli,
1971; Nardi, 1984), but the same as A. pallida and that
of most of the species belonging to this group (Nardi,
1984; 1988; 1989; 1991; Nardi & Nesi Nardi, 1987;
Ball & al., 1993; Constantinidis & al., 1997).
All the samples then collected by Nardi seemed to
show a globous hypocotyledonary tuberous root-
stock, in accordance with the main morphological fea-
tures of the group; the alternative type, the elongated
one, on the contrary, being typical of the A. longa, or
A. paucinervis or A. fontanesii, group (Nardi, 1984).
All these characteristics (vegetative features, root-
stock form and chromosome number) convinced Nar-
di that the new taxon could be described as A. pallida
subsp. castellana Nardi (Nardi, 1988 and pers. comm.),
which then became the westernmost of the group and
the only one present in the Iberian Peninsula.
During later investigations (Costa, 2002), new col-
lections and observations were carried out on this Iber-
ian taxon, leading to somewhat different conclusions.
Materials and methods
Main diagnostic features measurements of the leaf
and flower (Table 1, upper part) were taken from
about 20 different herbarium specimens belonging to
MA, MACB, MAF and FI (see Appendix).
Special care had been taken when choosing the
leaves and the corresponding axillary flower sampled:
only completely developed leaves, from the central
part of the stem, had been taken into account, avoid-
ing precocious ones (always of reduced size) and also
those too young, from the upper part of the stem.
The presence of an inferior, narrow ovary makes
difficult the recognition of the peduncle ends. To
overcome this problem, we have decided to take into
account the sum of the peduncle plus the ovary for the
comparisons.
All the new cited measurements of the tuberous
rootstock (Table 1, lower part) correspond to living
plants from the Alto del Mirlo (Madrid and Ávila,
Sierra de Gredos), cultivated in pots for a few years;
eventually, these plants unfortunately died.
Taxonomic teatment
From a morphological point of view, the vegetative
features of the original description have been con-
firmed, apart from a general slightly wider range of
variation (Table 1, upper part).
Surprisingly, however, the subterranean tuberous
rootstock has proven to belong to the elongated type:
in young individuals the general shape seems closer to
the globous type, but starts elongating in a little older
ones (Fig. 1), reaching a notable length in old plants
(Table 1, lower part).
All the samples of the holo- and isotypus (Nardi,
1988) are clearly young individuals, which fail to offer
a proper representation of the rootstock characte-
ristics.
A. Costa174
Anales del Jardín Botánico de Madrid 65(2): 173-178, julio-diciembre 2008. ISSN: 0211-1322
Fig. 1. An individual of Aristolochia castellana from Ávila, Casi-
llas. Picture by the author.
Table 1. Comparison of the main features measurements of the Aristolochia castellana with the protolog (Nardi, 1988). In the lower
part are some examples of the size of the tuberous rootstock from samples collected in the Alto del Mirlo (Madrid and Ávila, Sierra de
Gredos). In mm; l, length; w, width; bw, width at the base.
In Table 2 the morphological features of A. pallida
subsp. castellana are compared with the apparently
more related species, such as A. pallida subsp. pallida,
A. lutea and A. paucinervis Pomel, the latter always
found growing intermixed in the same localities (the
data for these other species come from another work
in preparation by the author).
As opposed to A. pallida subsp. pallida, A. pallida
subsp. castellana shares the chromosome number (2n
= 10), like other species belonging to the same group,
and some morphological characteristics, such as the
leaf shape and the petiole/(peduncle + ovary) ratio.
But other features are different: the flower limb is
bigger in A. pallida subsp. pallida and the extremely
important diagnostic character of the flower
limb/tube length ratio is significantly different, being
> 1 for A. pallida subsp. pallida, whilst < 1 for A. pal-
lida subsp. castellana. Moreover, the tuberous root-
stock shape, as was previously noted, belongs to the
elongated type in A. pallida subsp. castellana, whilst
belonging to the globous type in A. pallida subsp. pal-
lida.
The overall vegetative appearance of A. pallida sub-
sp. castellana is very similar to that of A. lutea, apart
from the generally reduced size of all parts and the
petiole/(peduncle + ovary) ratio, which is usually
higher in A. pallida subsp. castellana. Moreover, the
chromosome number (2n = 8 for A. lutea) and the
tuberous rootstock are different, the latter being
clearly globous in A. lutea.
The elongated tuberous rootstock type is the
Taxonomy of an Iberian Aristolochia
same as that of A. paucinervis, which is always found
growing intermixed with A. pallida subsp. castellana.
But A. paucinervis, which has a very much wider dis-
tribution (France, Iberian Peninsula and North
Africa) and is probably polyploid (2n = 36; Nardi,
1984), has significantly different vegetative character-
istics (Table 2): generally shorter petiole; a significant-
ly lower petiole/(peduncle + ovary) ratio; generally
bigger leaf blade with a more elongated shape; usual-
ly bigger flowers, with the tube proportionally wider
at the base.
All the above mentioned features caused us to
doubt as to the correct taxonomic treatment of this
new Iberian taxon, which would appear to have all the
characteristics for being considered an endemic
species proper.
Thus I propose the following new combination:
Aristolochia castellana (Nardi) Costa, comb. & stat.
nov. [Aristolochia pallida subsp. castellana Nardi,
Webbia, 42: 15, 16 fig. 1. 1988, basión. Holotypus:
Spain. Ávila: Cuevas del Valle, 24-V-1986, Nardi &
Nesi Nardi 8615 (FI; isotypus, MA 485799)].
Here below is a proposed modified dichotomous
key for the Iberian taxa of the genus Aristolochia
(compare with Castroviejo, 1986; Ball & al., 1993):
KEY OF ARISTOLOCHIA (ARISTOLOCHIACEAE)
1. Clustered flowers .......................................... A. clematitis
1. Solitary flowers ................................................................. 2
2. Climbing plant ................................................... A. baetica
2. Erect, ascending or prostrate plant ................................... 3
175
Anales del Jardín Botánico de Madrid 65(2): 173-178, julio-diciembre 2008. ISSN: 0211-1322
Table 2. Comparison of the morphological characteristics of Aristolochia castellana with some apparently related species. The data for
A. castellana correspond to 64 leaves from 26 stems; A. pallida 66 leaves / 30 stems; A. lutea 116 / 50; A. paucinervis 81 / 48. In mm;
mw, maximum width; the rest of legend as Table 1.
3. Sessile or subsessile, semi-amplexicaul leaves .... A. rotunda
3. Petiolate, never amplexicaul leaves ................................... 4
4. Fasciculate, rhizomatous woody roots, without a tuberous
rootstock; flower tube cylindrical, flower limb much wider
than tube and clearly revolute at the base; leaves densely
hairy on both surfaces, with a noticeable cartilaginous and
hairy margin, often undulate; main veins on the inferior sur-
face cartilaginous and hairy ......................... A. pistolochia
A. Costa
4. With a tuberous rootstock; flower tube obconic, club-
shaped, flower limb slightly wider than tube and slightly rev-
olute at the base; leaves glabrous to sparcely hairy, mainly on
the lower surface, with a glabrous, weakly cartilaginous, en-
tire margin; main veins on the inferior surface weakly carti-
laginous and glabrous or sparcely hairy ............................ 5
5. Generally prostrate stems; clearly elongated mature leaves,
16-35 × 5-20 mm, glabrous; flowers 10-30 mm long ..........
......................................................................... A. bianorii
5. Generally ascending stems; generally as long as wide or
slightly elongated mature leaves, 10-74 × 12-65 mm, weak-
ly haired, mainly on the lower surface; flowers 20-64 mm
long ................................................................................. 6
6. In completely developed leaves, petiole length more or less
equal to the relative flower peduncle + ovary length (Fig.
2a); perianth tube slightly narrower at the base, the maxi-
mum width usually around double that of the width at the
base (Table 2, mw / bw); leaves laminae usually slightly
elongated, almost glabrous on the upper surface .............
................................................................... A. paucinervis
6. In completely developed leaves, petiole clearly longer than
the relative flower peduncle + ovary (Fig. 2b); perianth tube
clearly narrower at the base, the maximum width usually
more than double that of the width at the base (Table 2, mw
/ bw); leaves laminae usually more or less as long as wide,
weakly hairy .................................................. A. castellana
Distribution and habitat
Similar to other endemic species of the genus in the
176
Anales del Jardín Botánico de Madrid 65(2): 173-178, julio-diciembre 2008. ISSN: 0211-1322
Fig. 2. Details of the main characteristic diagnostic features of:
a, Aristolochia paucinervis; b, A. castellana. Picture (a) by C. Bo-
tella from the Internet: Herbario Jaca, Gobierno de Aragón
(http://www.ipe.csic.es/floragon/index.php); picture (b) by the
author (the same individual as in Fig. 1).
Fig. 3. Distribution map of Aristolochia castellana.
a
b
Mediterranean region, this taxon distribution is fairly
reduced (Fig. 3), this is also probably due to its ecology.
We can usually find it on the wooded slopes of the
Sistema Central, usually in fresh and shady localities,
preferring slightly acid substrates, which is typical of
most of the Mediterranean species of this genus, asso-
ciated with Quercus spp. (generally Q. pyrenaica), Cas-
tanea sativa or rarely Pinus spp., always growing inter-
mixed with Aristolochia paucinervis.
It seems that its presence is correlated with the
mountain system and with a mother rock of acid plu-
tonic rock: its presently known distribution, based on
herbarium specimens and personal records without
voucher (see Appendix), strikingly coincides with the
presence of this type of rock in the Sistema Central:
Sierra de Gredos westward towards the Valle del Jerte
(Cáceres) and a small, isolated area in the Sierra de la
Peña de Francia (Salamanca; IGME, 1966; 1980).
Although this type of granite is also present in the
north of Portugal and Galicia (N-W Spain), to date,
no cited locality for this taxon in these areas is known.
Of course more investigations would be worthwhile,
but it may be that the climatic and ecological condi-
tions are not the most adequate for this taxon.
Also, no locality is yet known in the Sierra de
Guadarrama, the eastern part of the Sistema Central,
north-east of the Sierra de Gredos, but in this area the
mother rock is of a different type (IGME, 1966;
1980), which would seem to reinforce the idea of a
possible relation between this taxon and the afore-
mentioned geological characteristic, although exactly
what type of relation still remains to be studied.
Phylogenetic relationships
The different groups of the Mediterranean species
have been defined only on morphological bases (Nar-
di, 1984). A molecular phylogenetic study could re-
solve many questions, including the monophyly of the
proposed groups.
Nonetheless, the morphological and cytological
data described above could already help us to shed
light on this subject with regarding A. castellana.
The main morphological feature of the A. pallida
group is that of the petiole being very much longer
than the peduncle (Nardi, 1984). The most common
species of the group, A. lutea and A. pallida, both have
a globous tuberous rootstock, whilst the other few, all
narrow endemics (Nardi, 1984, 1989), have an elon-
gated one. The chromosome numbers are 2n = 8; 10,
all diploids, with the exception of A. tyrrhena Nardi &
Arrigoni (an endemic from Sardinia and Corsica; Nar-
di & Arrigoni, 1983), with 2n = 26 (Nardi, 1984) and
probably polyploid.
Taxonomy of an Iberian Aristolochia
Hence, from a morphological and cytological point
of view, A. castellana clearly belongs to this group.
The tuberous rootstock shape is worthy of com-
ment. It may depend on some ecological adaptation,
depending on the type and humidity of the substrate,
the elongated type being more frequent in dry and
stony soils and habitats. With regard to this question,
the case of A. insularis Nardi & Arrigoni (species be-
longing to the A. rotunda group; Nardi & Arrigoni,
1983; Nardi, 1984) is extremely interesting. De-
scribed at the beginning as a proper species, differing
from A. rotunda L. only in its elongated rootstock, it
was later changed to a subspecies of the latter, A. ro-
tunda subsp. insularis (Nardi & Arr.) Gamisans (Nar-
di, 1985; Nardi & Ricceri, 1987), after many interme-
diate individuals and populations had been found,
demonstrating, at least in this case, a sort of variability
between the two usually clearly distinguished shapes.
Coming back to A. castellana, we can consider it as
the only member of the A. pallida group present in the
Iberian Peninsula, the closest species, geographically,
of this group being A. pallida, whose westernmost dis-
tribution limit is the Rhone Valley in France (Nardi,
1984; Ball & al., 1993).
The A. longa group, whose most common member
is A. paucinervis, is mainly morphologically character-
ized by the elongated rootstock and the short size of
petiole and peduncle (Nardi, 1984). The chromosome
numbers are 2n = 12; 24; 36 (Nardi, 1984; Nardi &
Nesi Nardi, 1987), denoting the presence of poly-
ploidy, even if a proper polyploid series has not yet
been demonstrated.
Due to these characteristics, the relation of A.
castellana to this group appears unlikely.
All the other taxa of Aristolochia present in the
Iberian Peninsula are morphologically, cytologically,
geographically and ecologically very distant from
A. castellana.
Hybridization seems to be very rare in the genus,
with only very few cases known worldwide (Blanco,
2005); no hybrid has yet been described for the
Mediterranean species.
We might, therefore, describe A. castellana as a
diploid taxon; the westernmost and only one present
in the Iberian Peninsula at least distantly related to the
A. pallida group species; with a narrow, localized dis-
tribution, restricted to a single mountain system, geo-
logically well defined and isolated; spatially close to a
polyploid, morphologically different species.
These observations render a recent origin of this
taxon for speciation of a present Iberian species very
unlikely.
Thus we might consider A. castellana a relict, en-
177
Anales del Jardín Botánico de Madrid 65(2): 173-178, julio-diciembre 2008. ISSN: 0211-1322
demic (paleoendemic; Thompson, 2005), Iberian
species.
Acknowledgements
I am extremely grateful to Dr. Ginés López González (Real
Jardín Botánico, CSIC, Madrid, Spain), Dr. Nicolás López
Jiménez (Universidad Complutense, Madrid, Spain), Prof.
Christoph Neinhuis and Dr. Stefan Wanke (both of the Intitut für
Botanik, Technische Universität Dresden, Germany) for their
help, suggestions and fruitful conversations about the subject of
the present work.
A special thanks to Dr. Ramón Velasco Gemio, director of the
Reserva Natural de la Garganta de los Infiernos (Valle del Jerte,
Cáceres, Extremadura, Spain), for his kind collaboration, for giv-
ing permission to study the populations included in this protect-
ed area. Many thanks also to Dr. Ricardo Castroviejo Bolibar
(Universidad Politécnica de Madrid), for his kind explanations
and comments about the geological situation of the Sistema Cen-
tral.
Finally, I feel very grateful to the Editor-in-Chief and an Asso-
ciate Editor of this Journal, whose numerous, keen comments
have, I believe, increased, the quality of this small work.
References
Ball, P.W., Nardi, E. & Akeroyd, J.R. 1993. Aristolochia L. In:
Tutin, T.G., Burges, N.A., Chater, A.O., Edmondson, J.R., Hey-
wood, V.H., Moore, D.M., Valentine, D.H., Walters, S.M. &
Webb, D.A. (eds.), Flora Europaea 1: 87-89, 2nd ed., Cambridge.
Blanco, M.A. 2005. Un híbrido espontáneo entre Aristolochia
gorgona y A. grandiflora (Aristolochiaceae). Lankesteriana 5:
115-117.
Castroviejo, S. 1986. Aristolochiaceae. In: Castroviejo, S., Laínz,
M., López González, G., Monserrat, P., Muñoz Garmendia, F.,
Paiva, J. & Villar L. (eds.), Flora iberica 1: 201-205, Madrid.
Constantinidis, T., Kamari, G. & Phitos, D. 1997. A cytological
study of 28 phanerogams from the mountains of SE Sterea El-
las, Greece. Willdenowia 27: 121-142.
Costa, A. 2002. L’integrazione di tecniche morfologiche, citologiche
e molecolari nelle indagini di tassonomia vegetale: il caso di alcu-
ni taxa del gruppo Aristolochia pallida (Aristolochiaceae). Ph.D.
thesis. Università degli Studi di Firenze (Italy).
Fabbri, F. & Fagioli, A. 1971. Numeri cromosomici per la flora
italiana: 19-22. Informatore Botanico Italiano 3: 51-55.
IGME, Aranguren Sabas, F. (director) 1966. Mapa geológico de la
Península Ibérica, Baleares y Canarias; scale: 1:1,000,000; 5th
edition, Madrid.
IGME, Garcia-Loygorri, A. (director) 1980. Mapa geológico de la
Península Ibérica, Baleares y Canarias; scale: 1:1,000,000; 1st
edition, Madrid.
Nardi, E. 1984. The genus Aristolochia L. (Aristolochiaceae) in
Italy. Webbia 38: 221-300.
Nardi, E. 1985. Sulla tassonomia di Aristolochia insularis Nardi et
Arr. Webbia 39: 119-127.
Nardi, E. 1988. De Aristolochiae pallidae Willd. subspecie nova in
Hispania crescente. Webbia 42: 15-19.
A. Costa
Nardi, E. 1989. De speciebus Aristolochiae pallidae gregis (Aris-
tolochiaceae) in Graecia crescentibus. Willdenowia 18: 367-375.
Nardi, E. 1991. The genus Aristolochia L. (Aristolochiaceae) in
Greece. Webbia 45: 31-69.
Nardi, E. & Arrigoni, P.V. 1983. Le piante endemiche della Sardeg-
na: 134. Aristolochia tyrrhena Nardi et Arrigoni, species nova;
135. Aristolochia insularis Nardi et Arrigoni, species nova. Bol-
lettino della Società Sarda di Scienze Naturali 22: 347-356.
Nardi, E. & Nesi Nardi, C. 1987. Taxonomic and chorological notes
on the genus Aristolochia L. (Aristolochiaceae) from the Central
and Eastern Mediterranean area. Botanica Helvetica 97: 155-165.
Nardi, E. & Ricceri, C. 1987. Il genere Aristolochia L. in Corsica.
Webbia 41: 225-239.
Thompson, J.D. 2005. Plant Evolution in the Mediterranean: 40-
43, Oxford.
Appendix
Herbarium specimens
SPAIN. Ávila: El Barraco, Valle de Iruelas, Cañada del Perro,
5-V-1989, J.R. Molina s.n. (MACB 57804). Cuevas del Valle, 24-V-
1986, Nardi & Nesi Nardi 8615 (typus, FI; isotypus, MA 485799).
Hoyocasero, 5-VI-1992, 30TUK310728, F. Martínez García s.n.
(MACB 76808). Cáceres: La Garganta, 5-VI-1976, Fernández Diez
s.n. (MA 202936). Jerte,17-V-1995, M.A. Carrasco, J.M. Cardiel &
S. Pajarón s.n. (MACB 60461; MA 582305). Pto. Piornal, 4-V-
1990, M.A. Carrasco & S. Pajarón s.n. (MACB 35581; MA 504511).
Puerto de El Piornal, 23-V-1991, M.A. Carrasco, A. Escudero &
S. Pajarón s.n. (MACB 41363). Puerto del Piornal, vertiente N, 12-
V-1993, M.A. Carrasco, S. Pajarón & M.S. Borrado s.n. (MACB
49917; MA 529642). Valle del Jerte, castañar cerca del campamen-
to “Emperador Carlos”, 12-VI-1979, M.A. Carrasco & M. Velayos
s.n. (MACB 13131). Valle del Jerte, ribera del río, 18-IV-1975, Ca-
rrasco, Casaseca & Castroviejo s.n. (MACB 13127). Madrid: El Es-
corial, 18-VI-1992, 30TVK019956, F. Martínez García s.n. (MACB
76809). Salamanca: La Alberca, 14-V-1981, Fernández Diez s.n.
(MACB 8771). Toledo: Hinojosa de San Vicente, Monte San
Vicente, sobre el collado del Piélago, 2-VI-2002, 30TUK5344,
V.J. Arán 5173 & M.J. Tohá (MAF 163160). Sierra de San Vicente,
11-V-1993, M.A. Carrasco, S. Pajarón & M.S. Borrado s.n. (MACB
49897)
Populations visited by the author without voucher
SPAIN. Ávila: Alto del Mirlo, Casillas, 6-VI-2000, 30TUK66.
Barco de Ávila, 19-VI-2005, 30TTK86. Casillas, 19-VI-2005,
30TUK66. Cáceres: Hervás, 19-VI-2005, 30TTK56. Valle del
Jerte, Tornavacas, Dehesa La Campana, Paraje de Pedro Miguel,
18-VI-2005, 30TTK76. Valle del Jerte, Tornavacas, Reserva de la
Garganta de los Infiernos, Garganta de San Martín, la Casa Blan-
ca, 18-VI-2005, 30TTK75. Madrid: Alto del Mirlo, Rozas de Puer-
to Real, 6-VI-2000, 30TUK66. Salamanca: La Alberca, Las Batue-
cas, 19-VI-2005, 29TQE48. Toledo: Sierra de San Vicente, Nava-
morcuende, VI-2005, 30TUK54.
Associate Editor: J. Devesa
Received: 8-I-2008
Accepted: 5-VI-2008
178
Anales del Jardín Botánico de Madrid 65(2): 173-178, julio-diciembre 2008. ISSN: 0211-1322
