See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/342257610

The identity of Drimia purpurascens, with a new nomenclatural and

taxonomic approach to the “Drimia undata” group (Hyacinthaceae =
Asparagaceae subfam. Scilloideae)

Article  in  Plant Systematics and Evolution · August 2020

DOI: 10.1007/s00606-020-01689-1

CITATIONS READS

0 159

3 authors:

Manuel B. Crespo Mario Martínez Azorín

University of Alicante University of Alicante
510 PUBLICATIONS   3,572 CITATIONS    103 PUBLICATIONS   636 CITATIONS   

SEE PROFILE SEE PROFILE

Mª Ángeles Alonso-Vargas
University of Alicante
153 PUBLICATIONS   398 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Taxonomy and Syntaxonomy of halophilic species View project

Atlas de Flora Vascular Amenazada de España View project

All content following this page was uploaded by Manuel B. Crespo on 29 July 2020.

The user has requested enhancement of the downloaded file.

Plant Systematics and Evolution (2020) 306:67
https://doi.org/10.1007/s00606-020-01689-1

ORIGINAL ARTICLE

The identity of Drimia purpurascens, with a new nomenclatural

and taxonomic approach to the “Drimia undata” group
(Hyacinthaceae = Asparagaceae subfam. Scilloideae)
Manuel B. Crespo1   · Mario Martínez‑Azorín1   · Mª Ángeles Alonso1 

Received: 25 November 2019 / Accepted: 5 May 2020

© Springer-Verlag GmbH Austria, part of Springer Nature 2020

Abstract
The identity of Drimia purpurascens, a name usually synonymised to the African D. elata, is clarified. Morphological
data in the protologue allow identification of the former with the Mediterranean D. undata (Urginea undulata), which is
usually treated as belonging to Charybdis, a genus morphologically and molecularly separated from both Drimia s.str. and
Urginea. However, Charybdis is considered to be illegitimate and hence unavailable for use, and when treated as different
from Drimia and Urginea, the name Squilla might be applied. A binding decision was requested from the Nomenclature
Committee for Vascular Plants on whether Scilla and Squilla are sufficiently alike to be confused. In the meantime, Drimia
is here accepted to include, among others, two morphologically distinct aggregates: the “D. undata group” and the “D.
maritima group”. The present contribution focusses on the former, in which three species are recognised that show clear
morphological and biogeographical differences. Two specific names are revived: D. purpurascens (for D. undata) and D.
serotina (for D. tazensis), the latter established as a new combination. Furthermore, a new species (D. palaestina sp. nov.)
is described for the easternmost Mediterranean populations referred to D. undata. Similarly, one additional combination, D.
secundiflora, is effected in the “Drimia maritima group”. Nomenclatural types (including 3 new lectotypes, 2 new neotypes
and 1 new epitype) are indicated, and synonyms are shown for each accepted taxon. The taxonomic treatment for the “D.
undata group” including chromosome numbers, phenology, habitats, distribution, and conservation status is presented. An
identification key is also reported.

Keywords  Charybdis · Drimia · Mediterranean plants · Nomenclature · Taxonomy · Urginea

Introduction (Manning and Goldblatt 2018), albeit recent research sug-

gests it would contain at least 200 species (cf. Martínez-
When treated in a wide sense, Drimia Jacq. ex Willd. Azorín et al. 2019).
(Hyacinthaceae = Asparagaceae Urgineoideae) includes Molecular work undertaken in the last decades (cf. Man-
about 110 species distributed mostly throughout Africa ning et al. 2004; Pfosser and Speta 2001, 2004; Ali et al.
and extending into Eurasia and the Indian subcontinent 2013) revealed that most Mediterranean representatives
of Drimia (sensu lato) form a monophyletic aggregate, the
so-called Charybdis clade, also treated at generic rank as
Handling editor: Mike Thiv. Charybdis Speta. Members of this aggregate show a dis-
tinct morphology and biogeography, much diverging from
Electronic supplementary material  The online version of this the typical Drimia (sensu stricto).
article (https​://doi.org/10.1007/s0060​6-020-01689​-1) contains
Generic circumscription of subfamily Urgineoideae
supplementary material, which is available to authorized users.
has been particularly controversial in recent decades (cf.
* Manuel B. Crespo Martínez-Azorín et al. 2019). Based on morphological and
crespo@ua.es molecular data, both synthetic (e.g., Manning et al. 2004;
1 Manning and Goldblatt 2018) and analytic (e.g., Speta
Departamento de Ciencias Ambientales y Recursos
Naturales (dCARN), Universidad de Alicante, Apartado 99, 1998a, b, 2001; Pfosser and Speta 2001, 2004; Pfosser et al.
03080 Alicante, Spain 2012; Ali et al. 2013) arrangements have been adopted and

13
Vol.:(0123456789)
 67   Page 2 of 18 M. B. Crespo et al.

are currently available for use (see more detailed discussion in diameter at the base) and usually short (to 80 cm long);
in Crouch and Martínez-Azorín 2015; Pinter et al. 2013; or leaves narrow (to 3 cm wide), often strongly undulate on mar-
Martínez-Azorín et al. 2019). Recent Mediterranean floras gins, with minutely papillate to slightly long ciliate-papillate
monographs and/or databases accept Charybdis Speta at edges; inflorescence few-flowered (to 25 flowers), short and
generic rank (cf. Conti et al. 2005; Bacchetta et al. 2012; loose; capsule with 4−5 seeds per locule. On the one hand,
Véla et al. 2016; Bartolucci et al. 2018), whereas others D. undata [Scilla undulata Desf., Urginea undulata (Desf.)
include it in a widely circumscribed Drimia (cf. Dobignard Steinh., Charybdis undulata (Desf.) Speta] is said to occur
and Chatelain 2010; Tison and Foucault 2014; The Plant throughout the range of the group, mostly on coastal areas,
List 2020) or Urginea (cf. Valdés et al. 2002; Rico 2013). and it shows a rather broad morphological variation. On the
However, as remarked by Martínez-Azorín and Crespo other hand, D. tazensis (Batt. & Maire) Stearn [Urginea tazen-
(2016) Charybdis can be considered illegitimate under Art. sis (Batt. & Maire) Maire, Charybdis tazensis (Batt. & Maire)
52 of the Shenzhen Code (ICN, Turland et al. 2018) as it was Speta] is a W Moroccan endemic mostly distributed along
superfluous when published, a fact that makes it unavailable the Atlantic coast, the Middle Atlas and Anti-Atlas ranges
for use. Those authors have suggested Squilla Steinh. as an (cf. Rankou et al. 2015), and it is rather homogeneous in
available correct name at genus rank for that Mediterranean morphology.
group, though the possibility that Squilla and Scilla L. are Nonetheless, a detailed nomenclatural and morphological
sufficiently alike to be confused has led them to request for revision of the “Drimia undata group” reveals that the name
a binding decision on that matter to the Nomenclature Com- Drimia purpurascens is clearly connected with this latter
mittee for Vascular Plants, which still remains unresolved. In aggregate. Jacquin (1812) published his Drimia purpurascens
the meantime, we accept here an expanded Drimia (s.l.) to J.Jacq. for plants cultivated in Vienna (Austria), which were
accommodate provisionally members of that Mediterranean received from the Botanic Garden of Milan (Italy) under the
aggregate based on the nomenclatural issue explained above, name “D undulata”. The origin of those plants was unknown,
albeit we favour an analytic treatment in Urgineoideae (Mar- but they were placed at the Botanic Garden of Vienna together
tínez-Azorín et al. in prep.). with South African bulbous plants, and were implicitly
The “Charybdis clade” includes about nine to twelve taxa assumed to be native from southern Africa.
(cf. Maire 1958; Speta 1998a, 2001; Manning et al. 2004; From that time, Drimia purpurascens is widely accepted
Euro+Med 2020) that had been treated in the last century to be a synonym of D. elata Jacq. (e.g. Germishuizen and
in quite contrasting ranks (species, subspecies, varieties or Meyer 2003; Darbyshire et al. 2015; Manning and Goldblatt
forms) by different authors (see synonymies in The Plant 2018; The Plant List 2020), a species widespread from Sudan
List 2020). However, they can be easily arranged into two to South Africa. However, a close examination of the illus-
distinct morphological groups based on their floral charac- tration in the protologue of D. purpurascens (the only extant
teristics. Firstly, the “D. maritima group” is usually accepted original material of that name) reveals important morphologi-
to include about six species distinguished by their flowers cal differences that make synonymisation of both latter names
with tepals spreading, whitish with the midrib greenish unwise. Conversely, the main morphological characteristics of
to purplish; and the ovary about as long as the style, this D. purpurascens closely relate it to Drimia undata.
remaining hidden by the withered perigone after anthesis. In the present contribution we report nomenclatural and
They are widely distributed from the Canary Islands through taxonomic results on taxa of the “Drimia undata group”. We
S Europe and N Africa to the Middle East, being D. mar- show that three species can be easily recognised with clear
itima (L.) Stearn its most representative member. Secondly, morphological and biogeographical patterns. Two species
the “Drimia undata group” is often accepted to include two names are revived: D. purpurascens (for the well-known D.
species showing flowers with tepals recurved, pinkish to undata) and D. serotina (for D. tazensis), and a new species
olive-greenish with the midrib dark purple, and the ovary (D. palaestina) is described for the Eastern Mediterranean
much shorter than the style, this overtopping the withered populations usually referred to as D. undata. Nomenclatural
perigone after anthesis. They are found from Morocco and types (2 lectotypes, 2 neotypes and 1 epitype) are designated,
SE Iberian Peninsula through N Africa, Sicilia and Sardinia and the most relevant synonyms are shown for each accepted
to Palestina, being D. undata Stearn the most representative name. Furthermore, data on diagnostic morphological traits,
member. distribution areas (including maps), chromosome numbers,
Both species in the “D. undata group” (D. undata and and ecology are reported for the three accepted taxa. An iden-
D. tazensis) are also easy to distinguish from other rela- tification key is also presented which will favour future work
tives based on features such as the stems slender (to 3 mm on the group.

13
A new approach to the “Drimia undata” group Page 3 of 18  67

Materials and methods performed for 1.0 × 107 generations and sampled every

1000 generations. Two runs were executed. The general
Morphological, nomenclatural and habitat studies time reversible (GTR) + proportion of invariant sites
(I) + gamma distribution (G) model was used in the analy-
Detailed morphological studies were undertaken on wild ses (set nst = 6 rates = invgamma). The first 25% genera-
specimens and herbarium material following the terminol- tions (burninfrac = 0.25) were excluded, and the remaining
ogy used for species of Hyacinthaceae in Martínez-Azorín trees were used to compile a posterior probability (PP)
et al. (2007, 2009). Specimens and/or images from the her- distribution using a 50% majority-rule consensus. Collec-
baria ABH, BC, BCN, BM, C, CAIM, E, FI, GRA, HBG, tion data of the studied material are shown in Table 1 (the
HUJ, K, MA, MPU, P, VAL and W (acronyms according to Genbank accession numbers are not shown here, and will
Thiers 2020) were studied (Online Resource 1). Authors of be included in the ongoing phylogeny of the Urgineoideae
the cited taxa follow IPNI (2020). Typification of the con- by Martínez-Azorín et al. in prep.).
cerned names accords with the ICN (Turland et al. 2018).
The designation of the corresponding types is based on
consultation of the original material and/or the illustra- Results and discussion
tions cited in the respective protologues. The conservation
status is based on the IUCN (2012) Red List categories The identity of Drimia purpurascens
and criteria.
Joseph Jacquin (1812) described D. purpurascens based on
Preliminary phylogenetic studies material grown in the gardens of the University of Vienna,
without provenance, though supposedly coming from the
In the framework of an ongoing revision of Urgineoideae Cape of Good Hope. The description and illustration in the
(Martínez-Azorín et al. in prep.), a data matrix of 293 sam- protologue are detailed enough to get a complete picture
ples covering all major groups in the subfamily was con- of the vegetative and floral diagnostic characters of Jac-
structed, including concatenated trnL-trnF, matK and ycf quin’s plant (Fig. 1a). It is a glabrous plant exhibiting acute,
cpDNA nucleotide sequences for each sampled taxon. A linear lanceolate leaves, undulate on margins; bulb scales
preliminary Maximum Likelihood (ML) analysis was con- tightly arranged, whitish to greenish; bracts at anthesis much
ducted for 23 samples belonging to taxa in the “Charybdis shorter than pedicels, not comose; bracts small, inconspicu-
clade” (sensu Pfosser and Speta 1999, 2004), which com- ously spurred; flowers erect-patent; tepals connate at base to
prises some members of both the “Drimia undata group” form a very short tube and strongly reflexed free segments;
plus the “D. maritima” group, together with Drimia anom- stamens arising from the base of the perigone tube and never
ala (Baker) Baker (“Geschollia clade”), D. elata Jacq. connivent to the style, with yellowish anthers; and style as
(“Drimia s.str. clade”), D. karooica (Oberm.) J.C.Manning long as stamens.
& Goldblatt (“Rhadamanthopsis clade”), D. macrocentra The name D. purpurascens is usually regarded as con-
(Baker) Jessop (“Boosia clade”), D. marginata (Thunb.) specific with D. elata Jacq. (e.g. Germishuizen and Meyer
Jessop (“Austronea clade”) and D. physodes (Jacq.) Jes- 2003; Darbyshire et al. 2015; Manning and Goldblatt 2018;
sop (“Fusifilum clade”) as outgroup (Table 1), using the The Plant List 2020), an African plant widespread through
Tamura 3-parameter + gamma distribution (G) model after Sudan to South Africa (see http://powo.scien​ce.kew.org/taxon​
best fit model selection, as implemented in MEGA X (v. /urn:lsid:ipni.org:names​:53455​6-1), which slightly resem-
10.0.5; Kumar et al. 2018). The initial tree for the heu- bles Jacquin’s plant. However, the typical forms of D. elata
ristic search with Nearest-Neighbour-Interchange (NNI) strongly differ by its bulb scales more loosely imbricate, swol-
method was obtained by applying the Neighbour-Join and len and often reddish; bracts at anthesis equalling to longer
BioNJ algorithms to a matrix of pairwise distances esti- than pedicels and comose at the apex of the raceme; spur of
mated using the Maximum Composite Likelihood (MCL) bract to 2 mm long, clearly visible, 1/3–2/3 as long as bract;
approach. Support was assessed by bootstrap (Felsenstein flowers patent to slightly nodding; tepals connate at base to
1985) with 10,000 replicates, but holding only 10 trees per form a distinct tube and strongly reflexed free portion of seg-
replicate. Clades showing bootstrap support percentages ments; stamens arising from the upper part of the perigone
(BS) of 50–74% were considered as weakly supported, tube and all connivent to the style, with anthers ca 2 mm long,
75–89% moderately supported and 90–100% strongly sup- dark-coloured). Therefore, synonymisation of D. elata and
ported. Furthermore, a Bayesian inference (BI) analysis D. purpurascens is not warranted. Moreover, no plant similar
was conducted with MrBayes 3.2 (Ronquist et al. 2012). to the illustration in Jacquin’s (1812) protologue (Fig. 1a) is
The Markov chain Monte Carlo (MCMC) runs were known to occur in either the Cape flora or southern Africa.

13
 67   Page 4 of 18 M. B. Crespo et al.

Table 1  Collection data of samples used in the phylogenetic analyses of Drimia taxa from the “Charybdis” clade
Sample code Taxon Country Locality Collector(-s) Coll. Number Herbarium voucher
(ISO
code)

H822 D. aff. hesperia (Webb & MAR Ibn Qarrich-Zinad, T. Raus 286-07-95-20 –
Berthel.) J.C.Manning between Tetuan and
& Goldblatt Chauen
M1957 D. aff. hesperia MAR Tánger, 3 km antes del M. Martínez-Azorín et al. MMA1957 ABH80477
Cabo Espartel
M1959 D. aff. hesperia MAR Krimda, ca 18 km al NE M. Martínez-Azorín et al. MMA1959 ABH80480
de Larache
M1962 D. aff. hesperia MAR Larache, pinar al SW de M. Martínez-Azorín et al. MMA1962 ABH80481
la población
M1963 D. aff. hesperia MAR Dchar Abiad, entre M. Martínez-Azorín et al. MMA1963 ABH80483
Tánger y Tetuán
M1965 D. aff. hesperia MAR Tetuán, 6 km al sur, cerca M. Martínez-Azorín et al. MMA1965 ABH80486
de Dar Ben Karrich
H319 D. maritima (L.) Stearn ESP Prov. Madrid, central F. Speta s.n. W
Spain
H868 D. numidica (Jord. & ESP Balearic Islands: Ibiza, M. Pfosser and F. Speta Ibiza 15 –
Fourr.) J.C.Manning & close to Portinatx
Goldblatt beside main road
ABH80167 D. palaestina ISR Lakish Regional Council: A. Yoffe s.n. ABH80167
M.B.Crespo, Mart.- Nahal Adorayim, about
Azorín & M.Á.Alonso 4 km SW of Bnei
Dekalim
ABH80168 D. palaestina ISR Lakish Regional Council: A. Yoffe s.n. ABH80168
Nahal Adorayim, about
4 km SW of Bnei
Dekalim
H759 D. pancration (Steinh.) ITA Calabria, Frascineto F. Grims s.n. –
J.C.Manning & Gold-
blatt
ABH70437 D. pancration MLT Gozo island (Għawdex), J.L. Villar s.n. ABH70437
pr. Ir-Ramla, between
San Blas and S. Philip
M1942 D. secundiflora (Maire) MAR Kenitra: sand dunes, S of M.B. Crespo et al. MMA1942 ABH80153
M.B.Crespo, Mart.- Mehdia
Azorín & M.Á.Alonso
MOR1 D. secundiflora MAR Safi, 7 km N Safi on M.B. Crespo and M.Á. s.n. ABH79937
R301 to Ait Torkin Alonso
MOR2 D. secundiflora MAR Safi, 7 km N Safi on M.B. Crespo and M.Á. s.n. –
R301 to Ait Torkin Alonso
MOR3 D. secundiflora MAR Safi, 7 km N Safi on M.B. Crespo and M.Á. s.n. –
R301 to Ait Torkin Alonso
M1947 D. serotina (Schousb.) MAR Safi, 7.1 km NW on R301 M.B. Crespo et al. MMA1947 ABH80157
M.B.Crespo, Mart.- to Ait Torkin
Azorín & M.Á.Alonso
M1952 D. serotina MAR Safi, ca 20 km SW on M.B. Crespo et al. MMA1952 ABH80164
R301, ca 6 km SW of
Chaïbat
M1954 D. serotina MAR Tánger, Grutas de Hér- M. Martínez-Azorín et al. MMA1954 ABH80474
cules
M1961 D. serotina MAR Larache, pinar al SW de M. Martínez-Azorín et al. MMA1961 ABH80482
la población
ABH68248 D. purpurascens J.Jacq. ESP Valencia province: M. Martínez-Azorín s.n. ABH68248
Albaida, subida al pico
Benicadell desde el
Puerto de Albaida

13
A new approach to the “Drimia undata” group Page 5 of 18  67

Table 1  (continued)
Sample code Taxon Country Locality Collector(-s) Coll. Number Herbarium voucher
(ISO
code)

MA21519 D. purpurascens ALG Oran, Djebel Santo Ch. d’Alleizette s.n. MA21519
M1385 D. elata Jacq. ZAF E of Colenso, Zingela M. Martínez-Azorín et al. MMA1385 ABH74188
HBG03945 D. karooica (Oberm.) ZAF Nieuwoudtville, waterfall W. Wetschnig s.n HBG03945
J.C.Manning & Gold-
blatt
M1311 D. marginata (Thunb.) ZAF Calvinia, near Akkeren- M. Martínez-Azorín et al. MMA1311 ABH74881
Jessop dam N.R.
D16031 D. physodes (Jacq.) ZAF Aberdeen, Kariegasfon- A.P. Dold D16031 GRA​
Jessop tein
H840 D. macrocentra (Baker) ZAF E of Barkly Pass W. Wetschnig s.n. –
Jessop
H856 D. anomala (Baker) ZAF Grahamstown, hill W of W. Wetschnig s.n. HBG01510
Baker Caravan Park

Fig. 1  Drimia purpurascens. a Original illustration (lectotype, designated here) from Jacquin (1812). b Illustration of Scilla undulata from Des-
fontaines (1798), treated here in synonymy of the former

Conversely, Drimia undata Stearn, a plant distrib- Peninsula to Middle east), Corsica and Sardinia, looks very
uted through the S Mediterranean basin (from SE Iberian close to and exhibits vegetative and floral characteristics

13
 67   Page 6 of 18 M. B. Crespo et al.

fitting well with Jacquin’s D. purpurascens. It is worth commonly identified as U. undulata from N Africa, SE Ibe-
mentioning that the new name D. undata was proposed by rian Peninsula and Corsica show rather homogeneous fea-
Stearn (1978) to replace Urginea undulata (Scilla undulata, tures, including leaves green to somewhat glaucous, strongly
Charybdis undulata), since the binomial D. undulata Jacq. undulate on margins that are either hyaline and minute
was already in use for a quite different South African plant papillate or sometimes long ciliate-papillate (trichomes to
invalidly described (Art. 35.1 Ex. 1 of the ICN) by Nicolaus 0.25 mm long) on margins; scapes are often medium-sized
Jacquin (1797: 38) and validated later by Willdenow (1799). (to 60 cm tall); pedicels are 5–11 mm long, shorter than to
In that sense, his son (J. Jacquin 1812) mentioned in the about the length of perigone; tepals are typically purplish
protologue that bulbs of D. purpurascens were first received with darker midrib, about 3 mm wide and not markedly nar-
in the Botanic Garden of the University of Milan (Italy) rowed at base; fruiting pedicels are to 14 mm long, shorter
under the “erroneous name Drimia undulata”. This might to about equalling capsules; and seed are 6–8 × 3–4 mm
connect those bulbs to the Mediterranean plant, known by (Table 2). No geographic or ecologic patterns have been
that time as Scilla undulata, a name first published with a found linked to indumentum on leaf margin. All chromo-
short diagnosis by Desfontaines (1792) and later depicted some counts from Sardinia (Martinoli 1949), Spain (Valdés-
(Fig. 1b) and described in detail (Desfontaines 1798: 300, Bermejo 1980), Tunisia (Battaglia 1957) and Libya (Bartolo
tab. 88) by himself in his Flora atlantica (see for further et al. 1984) yielded a diploid set (2n = 2x = 20).
details Crespo et al. 2019). In our opinion, the plant illus- Regarding the Moroccan populations of the group, Maire
trated in the protologue of D. purpurascens fits perfectly (1931) first separated plants gathered near Taza as Urginea
with the current concept of Drimia undata (Fig. 1a, b), and undulata var. tazensis Batt. & Maire, which were charac-
therefore it is revived here to name the Mediterranean plant terised by distinctive flowers with perigone lobes narrower
at species rank in Drimia. Typification is effected here (see (ca 2 mm wide), markedly tapering at base (to 1 mm wide),
Taxonomic treatment section below) to stabilize further use greyish-green and lacking a darker central band; no data
of both names. on vegetative structures were cited. Later, the name U.
undulata var. major Gattef. & Weiller was applied to plants
Morphological variation in the “D. undata” group: growing in sandy soils of Tiznit and Safi (SW Morocco),
background which exhibited leaves longer and wider (to 30 × 1.5 cm)
than in the typical variety, and only slightly undulate on
Taxa related to Drimia purpurascens (the “D. undata margins (Gattefossé and Weiller 1938). No data on leaf edge
group” or also “U. undulata group”) have partially been characteristics were included, albeit close observation of
studied mainly in the W and C Mediterranean basin. Plants plants gathered in those territories (e.g. from Tiznit: MPU

Table 2  Comparison of the most important morphological features of the species related to Drimia purpurascens (the “D. undata aggregate”)
D. purpurascens D. serotina D. palaestina

Bulb (cm) 1.7–3.5 × 3–7 1.5–6.5 × 3–6.5 2–5 × 3–7

Leaves (cm) 7–15(–20) × (0.3–)0.5–1.5, mostly 20–30 × 1–1.5, erect, often barely 6–20 × 0.1–0.5, mostly appressed
appressed to substrate, strongly undulate on margins to substrate, strongly undulate on
undulate on margins margins
Leaf margin Hyaline, minutely papillate, sometimes Cartilaginous, long ciliate-papillate (to Cartilaginous, minutely papillate
long ciliate-papillate (to 0.25 mm) 0.25 mm) or minutely papillate
Stem (cm) (20–)25–40(–60) × (0.15–)0.25–0.3 60–80 × 0.2–0.4 (15–)20–40(–50) × 0.2–0.4
Bracts (mm) 2.8–5.5 2.4–4 2–3
Bracteoles 0.7–1.2 0–0.5 1–1.5
Flower pedicels (mm) 5–11 8–15 13–18
Tepals (mm) 9.5–12 × 2–3, linear-subspatulate, not 10–12 × 1.5–2, spatulate, markedly (11–)13–16 × 2–2.5, linear-subspatu-
markedly narrowed at base narrowed at base (to 1 mm wide) late, not markedly narrowed at base
Stamen filaments (mm) (4.5–)7–11 9–11 (10–)12–15
Anthers (mm) 1.5–3, oblong-linear 1–1.5(–2), subovate 1–1.5, subovate
Ovary Green Yellowish-green Green
Capsule (mm) 7–16 × 6–12 6–12 × 7–11 13.5–19 × 12–16
Fruiting pedicels (mm) 6–14, shorter to about equalling 9–12, equalling to 1.5 times longer (14–)20–35, 1.5–2 times longer than
capsule than capsule capsule
Seeds (mm) 6–8 × 3–4 6–7 × 3–4 (8.5–)10–12 × 5–7

13
A new approach to the “Drimia undata” group Page 7 of 18  67

barcodes MPU001417 [web!], MPU004613 [web!]; Safi: group. Among all studied material, the easternmost popula-
ABH barcodes ABH80158!, ABH80164!; Larache: ABH tions of the species, namely from Egypt (Cairo and North
barcode ABH80482!; Tánger: ABH barcodes ABH80473!, Sinai Governorates) to Israel and Jordan, show constant and
ABH80474!) exhibit broader leaves, mostly with smooth to consistent differences allowing easy identification (Table 2
weakly undulate and often slightly long ciliate-papillate mar- as D. palaestina). The oriental plants often produce narrow
gins. Other characters of all those Moroccan plants are simi- and glaucous leaves, with margin strongly undulate, carti-
lar to the typical U. undulata, excepting the purple-tinged laginous and minutely papillate, similar to those in the SW
base of leaves, often taller scapes (to 80 cm), inconspicuous Moroccan plants (Scilla serotina ≡ D. serotina); scapes are
or absent bracteoles (to 0.5 mm long), longer flower pedi- tall and many-flowered, also similar to those of D. serotina,
cels (to 15 mm long), smaller anthers (to 1.5 mm long) and but pedicels are constantly longer (to 18 mm long); flow-
fruiting pedicels (equalling to 1.5 times longer than capsule), ers show bigger tepals, (11–)13–16 × 2–2.5 mm, purplish to
which have diagnostic value when considered altogether olive-green suffused with purple and midrib dark purplish
(Table 2 as D. serotina). No caryological data appear to have to brownish, somewhat similar to the typical D. purpuras-
been reported for those plants so far. Both varieties above cens; fruiting pedicels are to 35 mm long, 1.5–2 times longer
were posteriorly synonymised and treated at specific rank, as than capsules, which are also bigger (13.5–19 × 12–16 mm);
U. tazensis (Batt. & Maire) Maire (cf. Maire 1940), Drimia finally, seeds are distinctly larger, reaching up to 12 × 7 mm.
tazensis (cf. Stearn 1978) or Charybdis tazensis (Batt. & Those latter reproductive characters are diagnostic for the
Maire) Speta (cf. Speta 2001), or also at subspecific rank, Palestinian plant and allow easy identification. Furthermore,
as Urginea undulata subsp. tazensis (Batt. & Maire) Maire Feinbrun-Dothan (1978) reported a tetraploid chromosome
& Weiller (cf. Maire 1958). Databases such as The Plant set (2n = 4x = 40) for those Palestinian plants, which places
List (2020), WCSP (2020) and Euro+Med (2020) or recent them apart from the remaining Mediterranean populations
N African floras (cf. Dobignard and Chatelain 2010) admit which have been reported as diploid (2n = 2x = 20). The
D. tazensis as the accepted name for that species. According combination of all those vegetative, reproductive and chro-
to the data shown above, and despite the scarce available mosomal characters is unique in the group (Table 2), and
herbarium material of this southern Moroccan plant, our support recognition of a new species which is described here
own studies on this biological entity lead us to favour the as D. palaestina (see Taxonomic treatment section below).
specific treatment.
However, the earliest valid available name for the Safi- Phylogenetic relationships in the “Drimia undata”
Tiznit-Taza species was proposed by Schousboe (1800), group
who described Scilla serotina Schousb. based on plants col-
lected near Safi. Although it has sometimes been regarded Our preliminary results on the phylogenetic relationships
as synonym with either U. anthericoides (Poir.) Steinh. (≡ within the Urgineoideae (Martínez-Azorín et al. in prep.)
Scilla anthericoides Poir. ≡ Charybdis anthericoides (Poir.) allow recognition of diverse strongly supported clades (not
Véla & Bélair) (cf. Baker 1873; Ball 1878) or the typical U. shown) which display a syndrome of morphological charac-
undulata (cf. Maire 1929), or even with D. maritima (L.) ters easy to recognise. One of those (the “Charybdis” clade
Stearn (cf. Govaerts 2011), the protologue of this name and sensu Pfosser and Speta 1999, 2004) corresponds to the spe-
the original material conserved in Copenhagen undoubtedly cies traditionally included in the groups of “Urginea mar-
point out to accept Scilla serotina as an earlier synonym of itima” (Drimia maritima) and “U. undulata” (D. undata),
U. tazensis. Consequently, the necessary new combination often segregated as the genus Charybdis, but provisionally
is proposed here in Drimia (see the Taxonomic treatment included here in a widely circumscribed Drimia.
section below). The final dataset for the “Charybdis” clade includes
The European populations of D. purpurascens (as U. 28 nucleotide sequences, with a total of 2665 positions.
undulata) are known to occur in Corsica, Sardinia and SE BI and ML methods yielded trees with similar topolo-
Iberian Peninsula (McNeill 1980; Rico 2013; Tison and gies. The obtained phylogenetic BI tree (Fig. 2) shows
Foucault 2014; Bartolucci et  al. 2018). This European a poor resolution in the basal relationships of the “Cha-
entity, sometimes considered as a different taxon at vari- rybdis clade”, probably because it includes only a partial
etal or subspecific rank (see the Taxonomic treatment sec- sampling of its diversity for which the analyses regions
tion below), should be now regarded as belonging to the D. are not discriminant enough. However, this clade forms a
purpurascens). strongly supported monophyletic group (1.0 PP, 100 BS)
Comparison of the main morphological features of liv- in which some trends are evidenced. Species belonging to
ing samples and the available herbarium vouchers from the the “D. maritima group” are arranged into three clades,
different known populations of D. purpurascens (the “D. each of them strongly supported. Firstly, the W Mediter-
undata group”) allows new insights into the taxonomy of the ranean taxa morphologically closer to D. maritima (D.

13
 67   Page 8 of 18 M. B. Crespo et al.

Fig. 2  Bayesian Inference (BI) phylogenetic tree for the “Charybdis” clade (sensu Pfosser and Speta 1999) of Drimia, estimated from the con-
catenated cpDNA (trnL-trnF, matK and ycf) data sets. Numbers above branches indicate Bayesian posterior probability (PP) values, and those
below branches correspond to bootstrap support percentages (BS)

maritima s. str., D. aff. hesperia and D. secundiflora) (1.00 PP, 94 BS). The genetic distances obtained within
form a strongly supported clade (1.00 PP, 99 BS), in samples of D. serotina range from 0.00036 to 0.00108,
which the internal relationships are not resolved. Sec- which are lower than distance values to both D. palaes-
ondly, two samples of the E-C Mediterranean D. pancra- tina (0.00180-0.00325) and D. purpurascens (0.00216 to
tion (1.00 PP, 99 BS) also show strong support and are 0.00325). Second, both samples of the newly described
also related to the sample of C Mediterranean D. numid- D. palaestina from the Eastern Mediterranean basin
ica. In a similar way, the three species belonging to the form a strongly supported clade (0.97 PP, 99 BS). The
“D. undata group” constitute clearly differentiated and genetic pairwise distance between them (0.00144) falls
strongly supported lineages, which exhibit noteworthy in the same range than values obtained in D. serotina,
genetic distances (Online resource 2) supporting treat- and the pairwise distance values of D. palaestina with
ment at species rank. First, all four samples of the W-S regard to D. purpurascens (0.00180–0.00288) are simi-
Moroccan D. serotina form a strongly supported clade lar to those obtained in D. serotina. Third, the samples

13
A new approach to the “Drimia undata” group Page 9 of 18  67

of D. purpurascens from Spain and Algeria also form a genus is illegitimate (since superfluous), as well as the com-
strongly supported clade (1.00 PP, 94 BS). The genetic binations effected under it
pairwise distance between both samples of D. purpuras- = Scilla undulata Desf., Méd. Eclairée Sci. Phys. 3(6):
cens (0.00036) is very low and also falls in the same rank 161. 1792, syn. nov. ≡ Epimenidion undulatum (Desf.)
as the values found within D. serotina and D. palaestina. Raf., Fl. Tellur. 2: 13. 1837 ≡ Prospero undulatum (Desf.)
It is worth mentioning that the genetic pairwise distances Salisb., Gen. Pl.: 28. 1866 ≡ Drimia undata Stearn, Ann.
found within and among taxa of the “D. undata group” Mus. Goulandris 4: 208. 1978, replac. name (non Drimia
are congruent with values obtained within and among undulata Jacq. ex Willd., Sp. Pl. ed. 4, 2(1): 166. 1799,
taxa of the “D. maritima group” which are traditionally nec Jacq., Collectanea 5 (Suppl.): 41. 1797, nom. inval.
treated as distinct species (Online resource 2). [Art. 35.1 Ex. 1 of the ICN]) ≡ Charybdis undulata
Although further samples of all three species covering (Desf.) Speta, Stapfia 75: 167. 2001, nom. illeg. [Art. 52
their whole distribution areas and additional nDNA nucle- of the ICN].—LECTOTYPE (designated by Crespo et al.
otide sequence data are needed, the preliminary molecular 2019): Desfontaines, Herbier de la Flore Atlantique, no.
relationships presented in Fig. 2 are congruent with the 707 Scilla undulata [ALGERIA. Tunis et Alger, colines
morphological affinities and divergences found among all arides, elle fleurit en hiver, 1787], Desfontaines (P bar-
those taxa, which together with the observed genetic dif- code P00320347!: the flowering raceme with bulb on the
ferences allow treating them at the specific rank. left side of the sheet), image available at: https​: //scien​
ce.mnhn.fr/insti ​ t utio ​ n /mnhn/colle ​ c tion ​ / p/item/p0032​
0347?listI​n dex=2&listC​o unt=6964; isolectotype: MPU
Taxonomic treatment barcode MPU017529 [web!]): only the flowering raceme
on the left side of the sheet, image available at: https​://
plant​s .jstor​. org/stabl​e /viewe​r /10.5555/al.ap.speci​m en.
According to the observed variation of taxa related to D. mpu01​7529
purpurascens (the “D. undata group”), at least three bio- = Urginea undulata f. caeculi Pau, Butl. Inst. Catalana
logical entities can be easily recognised on the basis of their Hist. Nat. 16: 136. 1916, syn. nov. ≡ Urginea undulata var.
morphological, molecular, cytological and geographic diver- caeculi (Pau) Pau, Bol. Soc. Ibér. Ci. Nat. 23: 91. 1924 ≡
gence (Table 2), for which the species rank is accepted here. Drimia undata subsp. caeculi (Pau) Mateo & M.B.Crespo,
Ongoing molecular and caryological work on the whole Fl. Abrev. Comun. Valenciana: 430. 1995 ≡ Urginea undu-
group will allow further clarification of the identity and rela- lata subsp. caeculi (Pau) M.B.Crespo & Mateo, Flora Mon-
tionships of some critical populations still poorly sampled. tiber. 5: 58. 1997.—LECTOTYPE (designated by Crespo
and Mateo 1997): SPAIN. Valencia prov., Alginet, cultivada,
Drimia purpurascens J.Jacq., Ecl. Pl. Rar. 1(3–4): 48, t. Aug 1916, F. Moroder (MA barcode MA21517!): the scape
30. 1812. Ind. loc.: “Das Vaterland ist zwar unbekannt, on the right side of the sheet, from a bulb with incipient
aber wahrscheinlich das Vorgebürge der guten Hoffnung.” leaves]
[Although its homeland is unknown, it comes probably Chromosome number: 2n = 2x = 20 (Martinoli 1949; Battag-
from the Promontory of Good Hope]. ≡ Idothea purpuras- lia 1957; Valdés Bermejo 1980; Bartolo et al. 1984).
cens (J.Jacq.) Kunth, Enum. Pl. 4: 342. 1843. ≡ Idothearia
purpurascens (J.Jacq.) C.Presl, Abh. Königl. Böhm. Ges. Phenology: Flowering from (June) July to October, fruiting
Wiss. ser. 5, 3: 544. 1845, comb. illeg. [Art. 52 of the from September to December; foliation from December to
ICN].—LECTOTYPE (designated here): [icon in] J.Jacq., May.
Ecl. Pl. Rar. 1(3–4), t. 30. 1812 (Fig. 1a): only the flower- Habitat: This species is usually found in clayish, chalky or
ing scape with bulb. EPITYPE (designated here): ITALY. sandy substrates, sometimes on rocky ground usually calcare-
Sardaigne, 1838, E. Thomas (P barcode P02166250!): the ous or decarbonised (“terra rossa”), often stony, on littoral
fragment on the left side of the sheet. Image available sunny slopes and flats, usually at low or medium elevations
at: http://media​photo​.mnhn.fr/media​/14413​08118​475Yu​ (0–800 m). It participates mostly in open Mediterranean
gdzUb​1wb0c​zBPy shrubland and steppe-grassland vegetation. Temperatures are
Note: The name Idothearia C.Presl was proposed to replace warm (thermomediterranean to mesomediterranean thermo-
Idothea Kunth, based on the existence of the earlier homo- climates) and yearly precipitations are moderate (between
nym name Idotea Fabricius, 1798 which applies to marine 100 and 500 mm, occasionally to 700 mm) with a peak in late
Isopoda (Crustacean Arthropods). However, homonymy autumn–winter (November–December), which mostly cor-
does not apply to organisms not governed by the Shenzhen respond to the (arid) semiarid to dry ombroclimates (sensu
Code [Princ. I and Art. 54 of the ICN], and hence Presl’s Rivas-Martínez 2005).

13
 67   Page 10 of 18 M. B. Crespo et al.

D. purpurascens
1035 km
D. serotina
D. palaestina htpps://google.es/maps

Fig. 3  Known distribution of Drimia purpurascens (circles), D. serotina (squares) and D. palaestina (triangles). Solid symbols indicate studied
specimens; open circles refer to bibliographic data. Satellite image from Google Earth Pro

Distribution area: SW Mediterranean basin (from W Algeria Some available herbarium material of this species is
to Libya), SE Spain, Sardinia, Corsica, and Alexandria (N mostly incomplete and does not allow full identification,
Egypt). References from Cyprus (cf. Meikle 1985: 1635) are or sometimes it belongs to wrongly identified specimens of
to be discarded (Fig. 3). It is the most widespread member Drimia fugax (Moris.) Stearn. New collections in the coastal
of the group. areas of NW Morocco will help to complete the distribution
area of the species and to elucidate its relationships with the
Additional specimens: See Online resource 1.
taxonomically close relative D. serotina (see below).
Conservation status: This species occupies the widest distri- For the Spanish populations of D. purpurascens, the
bution range in the group and the highest variety of habitats, name U. undulata f. caeculi Pau (≡ U. undulata var.
where no special threats seem to exist. Therefore, it should caeculi (Pau) Pau ≡ Drimia undata subsp. caeculi (Pau)
be labelled as “Least Concern” (LC). This assessment is Mateo & M.B.Crespo ≡ U. undulata subsp. caeculi (Pau)
coincident with that of Contu (2014) for the species consid- M.B.Crespo & Mateo) was proposed by Pau (1916), based
ered in a wider sense, including the Middle East populations on differences on flowering and fruiting features, as well
treated here as D. palaestina (see below). as on peculiarities of leaf margin (cf. Crespo and Mateo
1997; Rico 2013), and was said to differ from the North
Notes: No original material of D. purpurascens is currently
African populations at the subspecies rank (Mateo and
extant in Jacquin’s herbarium at W (Naturhistorisches
Crespo 1995; Crespo and Mateo 1997) and to be taxonom-
Museum Wien), since most vouchers were destroyed by fire
ically not different from the Corsican and Sardinian plants.
during the World War II (E. Vitek pers. comm.). Therefore,
Therefore, they are now regarded as belonging to the typi-
the illustration in the protologue is designated here as lecto-
cal D. purpurascens. The North African populations of
type of D. purpurascens. Furthermore, a voucher gathered
the species showing some slight differences with regard
in Sardinia (P barcode P02166250!) is also designated as
to European ones are the subject of current molecular and
epitype, since populations from the western Mediterranean
morphological research to clarify their taxonomic status.
basin (Corsica, Sardinia and the SE Iberian Peninsula) fit
It is worth mentioning that some coastal collections
best with morphological traits of the lectotype and proto-
of D. purpurascens made by A. Letourneux near Man-
logue. It is possible that the plants sent to Vienna from Milan
dara, NE of Alexandria (Egypt) in Oct 1878 (P barcode
were collected in Sardinia.
P01855114! p.p), Nov 1879 (P! barcode P01855115)

13
A new approach to the “Drimia undata” group Page 11 of 18  67

and Oct 1880 (P barcodes P01674855!, P01855113!, (designated here): Morocco, Tiznit, sandy soils, cultivated
P01855114! p.p., FI barcode FI054797 [web!]), and in in Alger (Algeria), 8 Sep 1936, R. Maire (MPU barcode
Lake Mariout close to Alexandria in Dec 1877 (K bar- MPU001417 [web!]). Image available at: https​: //herbi​
code K000483106!), include hysteranthous plants with er.umont​pelli​er.fr/zoomi​fy/zoomi​fy.php?fichi​er=MPU00​
leaves showing the margin minutely papillate; bulbs bear- 1417.
ing short, few-flowered scapes, flowers on short pedi- Note: No original material of that name appears to be avail-
cels (3–6 mm), bracts to 3.5 mm long and bracteoles ca able, and therefore a neotype is designated here. Probably,
0.6 mm long, tepals 9–11 mm long, anthers 0.9–1.1 mm the neotype corresponds to the same collection from Tiznit
long, and small capsules (7–11 mm wide) on short pedun- cited in the protologue, according to the gathering date.
cles (6–10 mm) and small seeds (5–7.5 × 3–5 mm long).
Chromosome number: Apparently not studied yet.
Although they grow close to the distribution area of D.
palaestina, the cited morphological characters fit per- Phenology: Flowering from August to October, fruiting from
fectly with D. purpurascens. They are the easternmost September to November; foliation from December to May.
populations of the species together with those in Libya (cf.
Viviani 1824; Durand and Barratte 1910; El-Gadi 1978). Habitat: This species is usually found in sandy substrates,
Unfortunately, no recent gatherings appear to exist from sometimes stony, on sunny slopes and flats, usually at low
Mandara and Alexandria bringing additional data. Further elevations (0–300 m). It participates mostly in open Mediter-
studies, including molecular work and cultivation of living ranean steppe-grassland and shrubland vegetation. Tempera-
plants, are being undertaken to clarify their taxonomic sta- tures are warm (thermomediterranean thermoclimate) and
tus and relationships, and for a final infraspecific arrange- yearly precipitations are scarce (between 100 and 550 mm)
ment of the whole group of D. purpurascens. with a main peak in late autumn-early winter (November–
December) and a secondary peak in late winter (February–
Drimia serotina (Schousb.) M.B.Crespo, Mar t.- March), which mostly correspond to the (hyperarid) arid to
Azorín & M.Á.Alonso, comb. nov. ≡ Scilla serotina dry ombroclimates (sensu Rivas-Martínez 2005).
Schousb., Iagttag. Vextrig. Marokko: 165. 1800, bas-
Distribution area: C, W and S Morocco, mainly in the Mid-
ionym. Ind. loc.: “Vulgaris in campis arenosis juxta
dle Atlas and Anti-Atlas ranges, extending northwards to
Saffy [Morocco]”.  ≡  Urginea anthericoides var. sero-
Taza and Melilla (cf. González García et al. 2003: 324, as
tina (Schousb.) Font Quer, Bol. Real Soc. Esp. Hist.
Urginea undulata), and getting rare southwards to the Saha-
Nat. 27: 45. 1927.—LECTOTYPE (designated here):
ran Atlas range (Fig. 3). It is the westernmost and southern-
[MOROCCO]. “Herb. Liebm[ann]. Marocco, Schousboe”
most member in the group of D. purpurascens.
(C barcode C10000939 [web!]) [IDC microfiche photo:
Type Herbarium nr. 183 I,4]. Image available at: https​:// Additional specimens: See Online resource 1.
plant​s.jstor​.org/stabl​e/10.5555/al.ap.speci​men.c1000​0939;
Conservation status: This species is endemic to Morocco,
isolectotype: C barcode C10000940 [web!]) [IDC micro-
where it occurs in a wide distributizzon range and a vari-
fiche photo: Type Herbarium nr. 183 I,5]
ety of habitats. However, only a few populations have been
= Urginea undulata var. tazensis Batt. & Maire, Bull. Soc.
confirmed and many threats seem to exist including land use
Hist. Nat. Afrique N. 22: 318. 1931, syn. nov. ≡ Urginea
change, large-scale farming, habitat fragmentation, human
tazensis (Batt. & Maire) Maire, Bull. Soc. Hist. Nat. Afrique
interference, infrastructure development, and urbanization.
N. 31: 43. 1940. ≡ Urginea undulata subsp. tazensis (Batt.
Therefore, according to Rankou et al. (2015) it should be
& Maire) Maire & Weiller, Fl. Afrique N. 5: 158. 1958. ≡
labelled as “Vulnerable” [VU, B2ab(ii,iii,v)]. However, new
Drimia tazensis (Batt. & Maire) Stearn, Ann. Mus. Gou-
prospections are still needed to more accurately map its dis-
landris 4: 208. 1978. ≡ Charybdis tazensis (Batt. & Maire)
tribution area and estimate the number of individuals per
Speta, Stapfia 75: 167. 2001.—NEOTYPE (designated
population.
here): MOROCCO. Tiznit, bulbs cultivated in Alger (Alge-
ria), 18 Sep 1939, R. Maire (MPU barcode MPU004613 Notes: The type material of Schousboe’s name conserved
[web!]). Image available at: https​://herbi​er.umont​pelli​er.fr/ in the Museum Botanicum Hauniense (lecto-: C barcode
zoomi​fy/zoomi​fy.php?fichi​er=MPU00​4613]. C10000939 [web!]; isolecto-: C barcode C10000940
Note: No original material of that name appears to be avail- [web!]), although lacking leaves, shows the typical flowers
able, and therefore a neotype is designated here from the of D. tazensis with bracteoles inconspicuous, tepals spatulate
material collected in the sandy plains N of Tiznit and culti- (ca 2 mm wide) strongly tapering at base (to ca 1 mm wide)
vated by Maire himself in Alger (cf. Maire 1940: 44) and shortly fused only at the very base, purplish-green, and
= Urginea undulata var. major Gattef. & Weiller, Bull. Soc. anthers about 1–1.3 mm long. Furthermore, it was described
Hist. Nat. Afrique N. 28: 539. 1938, syn. nov.—NEOTYPE from “Saffi”, a site where D. tazensis (= U. undulata var.

13
 67   Page 12 of 18 M. B. Crespo et al.

Fig. 4  Holotype specimen of
Drimia palaestina (HUJ No.
132724!)

major) is common and the typical D. purpurascens (D. the group of Drimia maritima (L.) Stearn. In fact, the veg-
undata) does not occur (cf. Maire 1958; Rankou et al. 2015; etative features of D. serotina resemble those of two both the
Fig. 3). Font Quer (1927) cited “Urginea anthericoides var. W Moroccan endemic D. secundiflora (Maire) M.B.Crespo
serotina” near Chiclana, Cádiz Province (southern Spain) & al. (see below) and the Algerian endemic D. anthericoides
to name some populations growing in coastal sandy soils. (Poir.) Véla & Bélair, two North African members of the “D.
Those Spanish populations, however, undoubtedly belong to maritima group”. However, the floral and fruit characters of

13
A new approach to the “Drimia undata” group Page 13 of 18  67

D. serotina allow unequivocal identification. It is worthy to or recurved on margins about the middle, obtuse to shortly
mention that in the protologue of Scilla serotina, Schous- mucronate, subcucullate and papillate at the apex, purplish
boe (1800) added two direct references to the diagnostic to olive-green suffused with purple, midrib dark purplish
phrase-names of Scilla anthericoides in Poiret (1789) and to brownish, slightly shrivelling after anthesis. Stamens
Desfontaines (1798), though in both cases with a question to 15 mm long, subequal or slightly longer than perigone;
mark. These facts did not make Schousboe’s name nomen- anthers greenish, 1.0–1.5 mm long, versatile; pollen yellow;
claturally superfluous when published and therefore it is a filaments purplish, subulate (flattened-winged when pressed,
legitimate name (Art. 52.2 Note 1 Ex. 11 of the ICN). Also, to 0.8 mm wide), (10–)12–15 mm long, about equalling the
it might justify the traditional (but erroneous) synonymisa- perigone lobes, erect-patent, spreading, sometimes slightly
tion of both latter names. curved one-side. Ovary 2–3.5 × 3–4  mm, subglobose to
ovoid, green; style purplish, (8–)10–12 mm long, sometimes
Drimia palaestina M.B.Crespo, Mart.-Azorín & slightly curved one side, persistent, longer than stamens and
M.Á.Alonso, sp. nov.—HOLOTYPE: Palestine, [Palestinian perigone after withering; stigma capitate, white, minutely
Authority], Judean Mts. South of Yata, Submediterranean papillate. Fruiting raceme subcylindrical, with peduncles
batha, 23 Sep 1970, Y. Plitmann and M. Raviv 26089 (HUJ (16–)20–35 mm long, mostly about twice longer than cap-
No. 132724!) (Fig. 4). sule, stout, erect-patent, incurvate at apex. Capsule trigo-
nous, globose, 13.5–19 × 12–16 mm, truncate at base, retuse
Etymology: The specific epithet “palaestina” refers to the at apex with an apiculum 0.5–1 mm long; valves shinning
historical geographical region of Palestina (Western Asia), on the inner surface, yellowish to somewhat nacarate. Seeds
where most populations of the species occur. elliptical-oblong, (8.5–)10–12 × 5–7 mm, subacute on one
Diagnosis: Species pulcherrima ad Drimiam purpuras- edge; testa black, shining, minutely alveolate. 2n = 4x = 40
centem affinis sed ea facile distinguenda praecipue foliis (Feinbrun-Dothan 1978).
margines minutissime papillosis cartilagineis; perigonii Phenology: Flowering from August to October, fruiting from
segmentis majoribus (11–)13–16 × 2–2.5 mm, cum pedicellis September to November; foliation from December to April.
12–18 mm long. per anthesin, quam eis plerumque aequan-
tibus vel paulo superantibus; staminum filamentis longiori- Habitat: This species is usually found in sandy, chalky, loess
bus; capsulis evidenter majoribus 14–19 × 12–16 mm, cum or clayish substrates, sometimes gravelly or stony, on sunny
fructiferis pedicellis (13–)20–35 mm long. quam eis saepis- hills and slopes, but occasionally enduring low salt contents
sime 1.5–2-plo longioribus; et seminibus valde majoribus in alluvial soils. It participates mostly in steppe-shrubland
8.5–12 × 5–7 mm. vegetation of desert and subdesert areas, but sometimes also
in Mediterranean shrubland vegetation in the northern parts
Description: Bulb ovoid, 2–5 × 3–7 cm; outer tunics papyra- of its distribution range (cf. Danin and Orshan 1999). Tem-
ceous and dark-greyish to brownish in colour, usually form- peratures are hot to warm (inframediterranean to thermo-
ing conspicuous zebrine remains at the apical part; the inner mediterranean thermoclimates) and yearly precipitations are
ones whitish to cream-coloured. Leaves 6–20 × 0.2–1.2 cm, scarce (between 90 and 300 mm, occasionally to 450 mm)
hysteranthous, numerous (5–16), usually glaucous-green- with a peak in winter (December–February), which mostly
ish, suberect to appressed to the ground, linear lanceolate, correspond to the hyperarid to semiarid (occasionally dry)
acute to subacute, strongly undulate with margin subcarti- ombroclimates (sensu Rivas-Martínez 2005).
laginous and minutely papillate, midrib prominent, slightly
Distribution: E Mediterranean basin (Palestine, Jordan,
canaliculated and curved downwards in the upper part. Stem
North Sinai and inland areas of NE Egypt between Cairo
(15–)20–40(–50) × 0.2–0.4 cm, erect, smooth, glaucous-
and Suez) (Fig. 3). It can be regarded as a Saharo-Arabian
greenish sometimes suffused with purple, pruinose. Raceme
chorologic element according to Danin and Plitmann (1985),
conic to subcylindrical, 15–28 cm long, with 8–50 flow-
which also penetrates into the neighbouring Mediterranean
ers loosely arranged. Pedicels 12–18 mm long at anthesis,
territories. It is the easternmost member of the group of D.
erect-patent to suberect, equalling to slightly longer than
purpurascens.
perigone, extending in fruiting plants to (16–)20–35 mm
long, each with one bract and bracteole, both usually early Additional specimens (paratypes): EGYPT. Cairo Govern.:
deciduous, whitish to pinkish; bracts 2–3 mm long, trian- Gebel Ghreibûn [29°51′00.0″N 31°45′00.0″E], 17 Nov 1884,
gular-acute, subulate, whitish to pinkish, with a basal spur G. Schweinifurth s.n. (K barcode K barcode K000483105!);
0.5 mm long; bracteoles 1–1.5 mm long, solitary, lateral, Gebel Abu Shama [29°52′00.0″N 31°38′00.0″E], 19 Apr
linear, whitish. Perigone 20–32 mm in diameter, stellate to 1924, N.D. Simpson 2769 (K barcodes K000483108!,
strongly reflexed, with lobes shortly connate at base up to K000483110!); ibidem, 23 Aug 1924, N.D. Simpson 2945
2 mm; tepals (11–)13–16 × 2–2.5 mm, linear-spatulate, flat (K barcode K000483107!); Cairo-Suez road, km 24, on

13
 67   Page 14 of 18 M. B. Crespo et al.

Fig. 5  Drimia purpurascens.
a Leaves, c flowers with short
pedicels, e capsules on short
pedicels, g seeds (ABH barcode
ABH19773!). Drimia palaes-
tina. b leaves, d flowers with
short pedicels, f capsules on
short pedicels, h seeds (HUJ
No. 132724!). Photographs:
José Quiles (Spain; http://www.
flora​silve​stre.es/medit​erran​ea/
index​.htm) for D. purpuras-
cens; Lior Almagor (Israel;
http://www.fromy​camer​a.com)
for D. palaestina; and Mario
Martínez-Azorín for seeds

hills of flint and sand with Pancratium sicken[bergeri]., 14 Northern Sinai, W El-Mizerie, sandy soil, 25 Apr 1959, L.
Jan 1945, P.H. Davis 8101 (E barcode E00345421 [web!]). Boulos 265 (CAIM [web!]). PALESTINE. [Jordan]. Moab,
North Sinai Govern.: Northern Sinai, El-Arish, cultivated 4 km S of El-Qatrani, alluvial gravelless soil, 750–800 m, 27
in Bot. Dept. Jerusalem, 1 May 1925, A. Eig 26082 (HUJ Mar 1936, A. Eig, M. Zohary and N. Feinbrun 26088 (HUJ
No. 132733!); ibidem, A. Eig 26083 (HUJ No. 132734!); No. 132731!). [Gaza] Wadi Sharia, on loess soils, 23 Nov

13
A new approach to the “Drimia undata” group Page 15 of 18  67

1942, P.H. Davis 4946 (K barcode K000483112!). [Israel]. D. purpurascens by its longer filaments of the stamens and
Northern Negev, Beer-Sheva to Tel-Melah, 4 Apr 1927, A. smaller bracts. Furthermore, the tetraploid chromosome set
Eig, M. Zohary and N. Feinbrun 26086 (HUJ No. 132719!); (2n = 40) of D. palaestina (Feinbrun-Dothan 1978) also sup-
Bershaeba [Beersheba], chalky hills, 22 Nov 1942, P.H. ports recognition at species rank from its diploid (2n = 20)
Davis 4963 (E barcode E00345422 [web!]); Sharon Plain, western Mediterranean relatives. Ongoing molecular work
env. of Wadi Faliq, cultivated in Mt. Scopus, Jerusalem, 15 on Urgineoideae will shed light on the more detailed phylo-
Sep 1931, N. Feinbrun 26084 (HUJ No. 132720!); Lakish genetic relationships of this outstanding new species.
Regional Council: Nahal Adorayim, about 4 km SW of Bnei
Dekalim, 31.500 N 34.880E, 300 m, on grey rendzina, 8 Identification key
Sep 2018, [ex Hort. Univ. Alicante, 9 Sep 2019 in flower],
A. Yoffe s.n. (ABH barcode ABH80167!); ibidem, [ex Hort. 1a. Perigone (11–)13–16 mm long; pedicels 12–18 mm
Univ. Alicante, 20 Oct 2019 in fruit, semiaborted], A. Yoffe long at anthesis, often longer than to equalling per-
s.n. (ABH barcode ABH 80168!); Northern Negev, Nahal igone; capsule 13.5–19 × 12–16 mm; fruiting pedicels
Gerar, near Tel-Sharia, gravel near water flow, 13 Nov 1998, (14–)20–35 mm long, often up to twice longer than cap-
A. Danin 72511 (HUJ No. 132729!); Northern Negev, hills sule; seeds (8.5–)10–12 × 5–7 mm ......... D. palaestina
near Mashabei-Sade, flint stones hill loess soil on chalks, 1 1b. Perigone 9.5–12 mm long; pedicels 5–11 mm long at
Sep 1995, O. Fragman 44521 (HUJ No. 132728!); Central anthesis, shorter than to equalling perigone; capsule
Negev, Makhtesh Hatira, 26 Jan 1944, T. Kushnir 19638 7–16 × 6–12  mm; fruiting pedicels 6–14  mm long,
(HUJ No. 132721!); Central Negev, Tel Yerocham to shorter than to slightly longer than capsule; seeds 6–8 × 
Dimona, 19 Sep 1965, G. Orhan 26085 (HUJ No. 132723!); 3–4 mm ........................................................................ 2
Central Negev, Negev Highlands, 18 km E of Dimona to 2a. Leaves mostly appressed to substrate, whitish to green-
Sedom, near Mtzad Tzafit, sandy ground, 23 Sep 1999, S. ish-white at base, without evident longitudinal furrows;
Einbinder and O. Steinitz 76812 (HUJ No. 132730!). tepals linear to oblanceolate, 2–3 mm broad, not or very
slightly tapering at base. Anthers (1.5–)2–3 mm long,
Conservation status: This species occupies a wide distribu-
oblong-linear at anthesis ..................... D. purpurascens
tion range and a variety of habitats, which perhaps might
2b. Leaves erect, purplish-red at base, often with 2 longitu-
suggest labelling it as “Least Concern” (LC). However, some
dinal furrows; tepals spatulate, 1.5–2 mm broad, taper-
potential threats seem to exist in vast territories of its range,
ing to 1 mm broad at base. Anthers 1–1.5(–2) mm long,
including land use change, large-scale farming, habitat frag-
subovate-oblong at anthesis ........................ D. serotina
mentation, human interference, water and soil pollution,
infrastructure development, and urbanization, which might
New combination in the “D. maritima group”
provoke rapid destruction of habitat leading to consideration
as “Vulnerable” (VU). Furthermore, new updated data are
needed regarding the number of extant populations, and pop- In the context of the taxonomic and molecular work here
ulation size and temporal dynamics in its entire distribution presented, the new combination below is needed for a name
area. Therefore, in the meantime it is here assessed as “Data in the “D. maritima group”, sometimes related to D. pur-
Deficient” (DD) pending a more accurate categorisation. purascens and its relatives:

Notes: Drimia palaestina shows a combination of charac- Drimia secundiflora (Maire) M.B.Crespo, Mart.-Azorín
ters not found in any of the remaining taxa in the group of & M.Á.Alonso, comb. nov. ≡ Urginea anthericoides var.
D. purpurascens (Table 2). The narrower strongly undulate secundiflora Maire in Bull. Soc. Hist. Nat. Afrique N. 24:
and glaucous leaves with margin cartilaginous and minutely 229. 1933, basionym. Ind. loc.: “Hab. in silva Mamora
papillate, as well as the bigger flowers, fruits and seeds are Imperii Maroccani occidentalis, solo arenoso”.—LECTO-
diagnostic features that allow easy recognition (Fig. 5). It TYPE (designated here): MOROCCO. Forêt de la Mamora,
is worth mentioning that a voucher exists at K including bulbes cultivés à Alger ayant fleuri en septembre 1932, 28
two collections that allow easy comparison of both cited Oct 1932, R. Maire s.n. (MPU barcode MPU 003124 [web!])
species, one (K barcode K000483106!) belonging to D. pur- [as “U. anthericoides var. secunda Maire” in sched.]): only
purascens from Mariout Lake near Alexandria, and the other the flowering scape. Image available at: https​://plant​s.jstor​
(K barcode K000483105!) belonging to D. palaestina from .org/stabl​e/viewe​r/10.5555/al.ap.speci​men.mpu00​3124.
Gebel Ghreibûn, near El Cairo. Drimia palaestina stands
closer to D. serotina on account of the taller stems, minutely Chromosome number: 2n = 6x = 60 (Battaglia 1964)
papillate margin of leaves, longer pedicels and greyish-
pinkish colour of flowers, but the latter exhibits broader and Distribution: W Morocco (from Larache and Forêt de la
less undulate, suberect leaves. The former also differs from Maâmora to Safi, and probably southwards).

13
 67   Page 16 of 18 M. B. Crespo et al.

Conservation status: Although this species is endemic to applications” (FlyHigh), the research project AICO-2019-003 from
western Morocco, many populations including numerous Generalitat Valenciana, and the grants ACIE18–03, UAUSTI18–02
and UAUSTI19-08 from University of Alicante. The authors declare
individuals occur throughout its distribution area in a variety that no competing interests exist.
of habitats, where no special threats seem to exist. Therefore,
it should be labelled as “Least Concern” (LC). Author contribution  M.B.C. conceived and planned the study, coor-
dinated tasks and wrote the major part of the text. M.M.A and M.Á.A.
Notes: This species was first restricted to the Maâmora For- produced DNA sequences, phylogenetic trees, photographs, maps and
est, close to Rabat. However, recent field work has demon- graphical artworks. All authors jointly gathered material and data in the
strated that it extends to the surroundings of Safi, and proba- field, conducted laboratory work, agreed on the final taxonomic treat-
ment, read the text, and significantly contributed to the final version.
bly southwards. It is a hexaploid (2n = 60) taxon characterised
by its smaller bulbs (to 8 cm diameter) with whitish-greenish
outer tunics; leaves smaller (to 30 × 3 cm) and shorter and Compliance with ethical standards 
slender scapes (to 80 × 0.7 cm); floral pedicels equalling or
Conflict of interest  The authors declare that they have no conflict of
a little longer than the perigone pieces; flowers with white interests.
tepals 10–12.5 mm long, showing a conspicuous reddish
midrib, stamen filaments 6–8 mm long, and yellowish ovary Information on Electronic Supplementary Mate-
4.5–5 mm long, oblong; fruits 13–16 mm long, unilateral, rial
obtuse to subemarginate, not apiculate or with an inconspicu-
ous apiculus to 0.6 mm long at the apex; and seeds 8–9.5 mm Online Resource 1. Additional studied material of the “Charybdis
clade” in Drimia s.l.
long. On the basis of the smaller bulbs, leaves and scapes,
Online Resource 2. Matrix of the pairwise distances obtained for the
D. secundiflora resembles D. anthericoides (Poir.) Véla & studied taxa of the “Charybdis clade” and outgroups in Drimia s.l.
Bélair, a diploid (2n = 20) species endemic to Algeria and
Tunisia, but the distinct characteristics of flowers, fruits and
seed in the latter (see Véla et al. 2016; Véla and Bélair 2019) References
allow easy differentiation. In some extent, the flower features
Ali SS, Pfosser M, Wetschnig W, Martínez-Azorín M, Crespo MB,
of D. secundiflora resemble those of D. maritima, a hexa- Yu Y (2013) Out of Africa: Miocene dispersal, vicariance, and
ploid taxon from the Iberian Peninsula, well-characterised by extinction within Hyacinthaceae subfamily Urgineoideae. J Integr
its robust, taller scapes; flowers more numerous, larger, with Pl Biol 55:950–964. https​://doi.org/10.1111/jipb.12065​
tepals to 14 mm long; stamen filaments to 10 mm long; and Bacchetta G, Brullo S, Demerico S, Pontecorvo C, Salmeri C (2012)
Charybdis glaucophylla (Asparagaceae), a A new species from
ovary 6–7 mm, greenish-yellow; fruits 12–14 mm long, not Sardinia. Phytotaxa 69:16–26. https​://doi.org/10.11646​/phyto​
unilateral, subacute, with a marked apiculus ca 1 mm long taxa.69.1.4
at the apex; and seeds 6–8 mm long. Therefore, based on the Baker JG (1873) Revision of the genera and species of Scilleae and
morphological, chorological and caryological divergences Chlorogaleae. J Linn Soc Bot 13:209–236
Ball J (1878) Spicilegium Florae Maroccanae. Part 2, containing
with regard to its closest relatives, Maire’s (1933) taxon is descriptions of genera and species, Malvaceae to Umbelliferae.
treated here at specific rank. Bot J Linn Soc 16:377–772
Bartolo G, Brullo S, Pavone P, Terrasi MC (1984) Cytotaxonomical
Acknowledgements  We thank the curators and staff of the herbaria notes on some Liliaceae of N. Cyrenaica. Webbia 38:601–622.
cited in the text, especially Hagar Leshner for the loan of material from https​://doi.org/10.1080/00837​792.1984.10670​329
the Middle East at HUJ. Cécile Aupic and Véronique Andro (Herbier Bartolucci F, Peruzzi L, Galasso G, Albano A, Alessandrini A,
P, Muséum National d’Histoire Naturelle, Paris) kindly helped, respec- Ardenghi NMG, Astuti G, Bacchetta G, Ballelli S, Banfi E,
tively, with Desfontaines’s historical herbarium and digital images of Barberis G, Bernardo L, Bouvet D, Bovio M, Cecchi L, DiPi-
Scilla undulata. Dr Leopoldo Medina (Real Jardín Botánico, Madrid) etro R, Domina G, Fascetti S, Fenu G, Festi F, Foggi B, Gallo
provided material of Drimia undulata for molecular work. Mr Atay L, Gottschlich G, Gubellini L, Iamonico D, Iberite M, Jiménez-
Yoffe (Kibbutz Nativ HaLamed-He, Israel) kindly provided bulbs and Mejías P, Lattanzi E, Marchetti D, Martinetto E, Masin RR, Meda-
seeds of Drimia palaestina. Juan A. González-García (Campus of gli P, Passalacqua NG, Peccenini S, Pennesi R, Pierini B, Poldini
Melilla, University of Granada) collected material of Drimia serotina L, Prosser F, Raimondo FM, Roma-Marzio F, Rosati L, Santan-
from Melilla. Norbert Holstein (British Museum Natural History, Lon- gelo A, Scoppola A, Scortegagna S, Selvaggi A, Selvi F, Soldano
don), and Harry Smith (Royal Botanic Gardens, Kew) forwarded digital A, Stinca A, Wagensommer RP, Wilhalm T, Conti F (2018) An
images of material at BM and K, respectively. Nael M. Fawzi (Agricul- updated checklist of the vascular flora native to Italy. Pl Biosyst
tural Research Center, Cairo) supplied digital images and information 152:179–303. https​://doi.org/10.1080/11263​504.2017.14199​96
on material from Egypt at CAIM. Also, José Quiles (Spain; http:// Battaglia E (1957) Filogenesi del cariotipo nel genere Urginea. I–
www.flora​silve​stre.es/) and Lior Almagor (Israel; http://www.fromy​ III: U. maritima (L.) Baker, U. fugax (Moris) Steinh. ed U.
camer​a.com) are kindly thanked for providing, respectively, photo- undulata (Desf.) Steinh. Caryologia 9:234–273. http://dx.doi.
graphs in habitat of D. purpurascens (D. undata) and D. palaestina. org/10.1080/00087​114.1957.10797​593
This work was partly supported by H2020 Research and Innovation Battaglia E (1964) Urginea maritima (L.) Baker: nuovi reperti di
Staff Exchange Programme of the European Commission, project biotipi cariologici 2n, 3n, 4n, 6n. Caryologia 17:509–518. https​
645636: “Insect-plant relationships: insights into biodiversity and new ://doi.org/10.1080/00087​114.1964.10796​147

13
A new approach to the “Drimia undata” group Page 17 of 18  67

Conti F, Abbate G, Alessandrini A, Blasi C (eds) (2005) An annotated Jacquin JF (1812) Eclogae plantarum rariorum aut minus cognitarum,
checklist of the Italian vascular Flora. Palombi Editori, Roma vol. 1(3–4). A. Strauss, Vindobonae [Wien]
Contu S (2014) Drimia undata. The IUCN Red List of Threatened Spe- Kumar S, Stecher G, Li M, Knyaz C, Tamura K (2018) MEGA X: Molecu-
cies 2014: e.T13148003A44453510. Available at: http://dx.doi. lar evolutionary genetics analysis across computing platforms. Molec
org/10.2305/IUCN.UK.2014-1.RLTS.T1314​8003A​44453​510.en. Biol Evol 35:1547–1549. https​://doi.org/10.1093/molbe​v/msy09​6
Accessed 8 Jun 2019 Maire R (1929) Contributions à l´étude de la flore de l’Afrique du Nord.
Crespo MB, Mateo G (1997) El género Urginea Steinh. (Hyacintha- Fascicule 16(1). Bull Soc Hist Nat Afrique N 20:171–220
ceae) en la flora ibérica. Flora Montiber 5:56–60 Maire R (1931) Contributions à l´étude de la flore de l’Afrique du Nord.
Crespo MB, Martínez-Azorín M, Alonso MÁ (2019) Lectotype des- Fascicule 18. Bull Soc Hist Nat Afrique N 22:275–325
ignation for Scilla undulata Desf. (Asparagaceae) from among Maire R (1933) Contributions à l´étude de la flore de l’Afrique du
the mixed original material. Phytotaxa 411:89–92. http://dx.doi. Nord. Fascicule 20. Bull Soc Hist Nat Afrique N 24:194–232
org/10.11646​/phyto​taxa.411.1.9 Maire R (1940) Contributions à l´étude de la flore de l’Afrique du
Crouch NR, Martínez-Azorín M (2015) Drimia edwardsii (Aspara- Nord. Fascicule 29. Bull Soc Hist Nat Afrique N 31:7–49
gaceae, Scilloideae), a new urgineoid species from the Mkhom- Maire R (1958) Flore de l’Afrique du Nord, vol 5. Lechevalier, Paris
azi River Valley of eastern South Africa. Phytotaxa 195:135– Manning JC, Goldblatt P (2018). Systematics of Drimia Jacq.
144. http://dx.doi.org/10.11646​/phyto​taxa.195.2.2 (Hyacinthaceae: Urgineoideae) in southern Africa. Strelitzia
Danin A, Orshan G (1999) Vegetation of Israel I. Desert and coastal 40:1–173
vegetation. Backhuys Publishers, Leiden Manning JC, Goldblatt P, Fay MF (2004) A revised generic synopsis
Danin A, Plitmann U (1985) Revision of plant geographical terri- of Hyacinthaceae in Sub-Saharan Africa, based on molecular
tories of Israel and Sinai. Pl Syst Evol 156:43–53. https​://doi. evidence, including new combinations and the new tribe Pseu-
org/10.1007/BF009​37200​ doprospereae. Edinburgh J Bot 60:533–568
Darbyshire I, Kordofani M, Farag I, Candiga R, Pickering H (eds) (2015) Martínez-Azorín M, Crespo MB (2016) (48) Request for a bind-
The Plants of Sudan and South Sudan. Kew publishing, Kew ing decision on whether Scilla L. (Hyacinthaceae subfam.
Desfontaines RL (1792) Décade des plantes nouvelles, dont les Hyacinthoideae) and Squilla Steinh. (Hyacinthaceae subfam.
graines ont été apportées des côtes de Barbarie par M. Desfon- Urgineoideae) are sufficiently alike to be confused. Taxon
taines, et qui sont maintenant propagées dans nos jardins. Méd 65:1437–1438. https​://doi.org/10.12705​/656.27
Eclairée Sci Phys 3:161–163 Martínez-Azorín M, Crespo MB, Juan A (2007) Taxonomic revision
Desfontaines RL (1798) Flora atlantica, sive historia plantarum quae of Ornithogalum subg. Cathissa (Salisb.) Baker (Hyacintha-
in Atlante, agro tunetano et algeriensi crescunt, vol. 1. Typis C. ceae). Anales Jard Bot Madrid 64:7–25
Panckoucke, Parisiis [Paris]. https​://doi.org/10.5962/bhl.title​.323 Martínez-Azorín M, Crespo MB, Juan A (2009) Taxonomic revision
Dobignard D, Chatelain C (2010) Index synonymique de la flore of Ornithogalum subg. Beryllis (Hyacinthaceae) in the Iberian
d’Afrique du nord, 1. Éditions des Conservatoire et Jardin bota- Peninsula and the Balearic Islands. Belgian J Bot 142:140–162.
niques, Genève https​://doi.org/10.2307/41427​183
Durand E, Barrate G (1910) Florae libycae prodromus, ou Catalogue Martínez-Azorín M, Crespo MB, Alonso MÁ, Dold AP, Crouch NR,
raisonné des plantes de Tripolitaine. Imprimerie Romet, Genève Pfosser M, Mucina L, Pinter M, Wetschnig W (2019) New combi-
El-Gadi A (1978) Urginea Steinh. In: Jafri SMH, El-Gadi A (eds) nations in the tribe Urgineeae (Asparagaceae subfam. Scilloideae)
Flora of Libya, vol. 57. Al Faateh University, Faculty of Science, with comments on contrasting taxonomic treatments. Phytotaxa
Department of Botany, Tripoli, pp 35–40 397:291–299. https​://doi.org/10.11646​/phyto​taxa.397.4.3
Euro + Med (2020) The Euro + Med plant base—the information Martinoli G (1949) Ricerche citotassonomiche sui generi Urginea
resource for Euro-Mediterranean plant diversity. Available e Scilla della flora sarda. Caryologia 1:329–357. https​://doi.
at: http://www.emplan​ tbase​ .org/home.html. Accessed 10 Jan 2020 org/10.1080/00087​114.1949.10797​517
Feinbrun-Dothan N (1978) Flora palaestina, vol. 3. Israel Academy of Mateo G, Crespo MB (1995) Flora abreviada de la Comunidad Valen-
Sciences and Humanities, Jerusalem ciana. Gamma, Alicante
Felsenstein J (1985) Confidence limits on phylogenies: an approach McNeill J (1980) Urginea Steinh. In: Tutin TG, Heywood VH, Burges
using the bootstrap. Evolution 39:783–791. https ​ : //doi. NA, Moore DM, Valentine DH, Walters SM, Webb DA (eds)
org/10.2307/24086​78 Flora europaea, vol. 5. Cambridge University Press, Cambridge,
Font Quer P (1927) Notas sobre la flora gaditana. Bol Soc Esp Hist pp 40–41
Nat 27:39–46 Meikle RD (1985) Flora of Cyprus, vol. 2. The Bentham-Moxon Trust,
Gattefossé J, Weiller M (1938) Contribution à la connaisance de la Kew
flore du Maroc. Fascicule I. Deuxieme partie. Bull Soc Hist Nat Pau C (1916) Urginea undulata (Desf.) Steinh. Especie nueva para
Afrique N 28:539–540 el continente europeo. Butll Inst Catalana Hist Nat 16:135–136
Germishuizen G, Meyer NL (eds) (2003) Plants of Southern Africa: Pfosser M, Speta F (1999) Phylogenetics of Hyacinthaceae based on
an annotated checklist. Strelitzia 14. National Botanical Institute, plastid DNA sequences. Ann Missouri Bot Gard 86:852–875.
Pretoria https​://doi.org/10.2307/26661​72
González García JA, García Peña H, Cabo Hernández JM (2003) La Pfosser M, Speta F (2001) Bufadienolides and DNA sequences: on
flora silvestre de Melilla. Consejería de Medio Ambiente, Melilla lumping and smashing of subfamily Urgineoideae (Hyacintha-
Govaerts RHA (2011) World checklist of selected plant families pub- ceae). Stapfia 75:177–250
lished update. Facilitated by the Trustees of the Royal Botanic Pfosser M, Speta F (2004) From Scilla to Charybdis—Is our voyage
Gardens, Kew safer now? Pl Syst Evol 246:245–263. https​://doi.org/10.1007/
IPNI (2020) [continuously updated]. The international plant names s0060​6-004-0153-z
index. Available at: http://www.ipni.org. Accessed 10 Jan 2020 Pfosser M, Knirsch W, Pinter M, Ali S, Dutta S, Wetschnig W (2012)
IUCN (2012) IUCN Red List categories and criteria: version 3.1, 2nd Phylogenetic relationships of Malagasy Hyacinthaceae. Pl Syst
edn. IUCN Species Survival Commission, Cambridge Evol 145:65–72. https​://doi.org/10.5091/plece​vo.2012.590
Jacquin NJ (1797) Collectanea ad botanicam, chemiam, et historiam Pinter M, Crespo MB, Ilg I, Luidold AK, Martínez-Azorín M, Müller-
naturalem spectantia, vol. 5 (Collectaneorum supplementum). Doblies U, Müller-Doblies D, Pfosser M, Wetschnig W (2013)
Officina Wappleriana, Vindobonae [Wien] Mucinaea (Hyacinthaceae-Urgineoideae), a remarkable new

13
 67   Page 18 of 18 M. B. Crespo et al.

genus from Namaqualand (Northern Cape Province, South Botanical Garden’s Virtual Herbarium. Available at: http://sweet​
Africa). Phyton (Horn) 53:289–304 gum.nybg.org/scien​ce/ih/. Accessed 10 Jan 2020
Poiret JLM (1789) Voyage en Barbarie, vol. 2. Libraire J. Née de la Tison J-M, de Foucault B (2014) Flora Gallica. Flore de France, Mèze
Rochelle, Paris Turland NJ, Wiersema JH, Barrie FR, Greuter W, Hawksworth DL,
Rankou H, Culham A, Taleb MS, Ouhammou A, Martin G, Jury SL Herendeen PS, Knapp S, Kusber W-H, Li D-Z, Marhold K, May
(2015) Conservation assessments and Red Listing of the endemic TW, McNeill J, Monro AM, Prado J, Price MJ & Smith GF (eds)
Moroccan flora (monocotyledons). Bot J Linn Soc 177:504–575. (2018) International Code of Nomenclature for algae, fungi, and
https​://doi.org/10.1111/boj.12258​ plants (Shenzhen Code) adopted by the Nineteenth International
Rico E (2013) Urginea Steinh. In: Rico E. Crespo MB, Quintanar Botanical Congress Shenzhen, China, July 2017. Regnum Veg-
A, Herrero A, Aedo C (eds) Flora iberica, vol. 10. Real Jardín etabile 159. Koeltz Scientific Books, Königstein. https​://doi.
Botánico, CSIC, Madrid, pp 210–217 org/10.12705​/Code.2018
Rivas-Martínez S (2005) Avances en Geobotánica. Real Academia Valdés B, Rejdali M, Achhal El Kadmir A, Jury SL, Montserrat JM
Nacional de Farmacia, Madrid (eds) (2002) Catalogue des plantes vasculaires du Nord du Maroc.
Ronquist F, Teslenko M, van der Mark P, Ayres DL, Darling A, Consejo Superior de Investigaciones Científicas, CSIC, Madrid
Höhna S, Larget B, Liu L, Suchard MA, Huelsenbeck JP (2012) Valdés-Bermejo E (1980) Números cromosomáticos de plantas occi-
MrBayes 3.2: efficient Bayesian phylogenetic inference and model dentales, 55–63. Anales Jard Bot Madrid 37:193–198
choice across a large model space. Syst Biol 61:1–4. https​://doi. Véla E, de Bélair G (2019) Charybdis anthericoides. Pp. 423–424 in:
org/10.1093/sysbi​o/sys02​9 Raab-Straube E. von, Raus Th (eds.), Euro + Med-Checklist Notu-
Schousboe PKA (1800) Iagttagelser over Vextriget i Marokko. Trykt lae, 11 [Notulae ad floram euro-mediterraneam pertinentes No. 40].
for Forfatteren hos K.H. Seidelin, Kjöbenhavn. https​://bibdi​gital​ Willdenowia 49:421–445. https​://doi.org/10.3372/wi.49.49312​
.rjb.csic.es/idurl​/1/11442​ Véla E, de Bélair G, Rosato M, Rosselló JA (2016) Taxonomic remarks
Speta F (1998a) Hyacinthaceae. In: Kubitzki K (ed) The families and on Scilla anthericoides Poir. (Asparagaceae, Scilloideae), a
genera of vascular plants, vol. 3. Springer, Berlin, pp 261–285 neglected species from Algeria. Phytotaxa 288:154–160. http://
Speta F (1998b) Systematische Analyse der Gattung Scilla L. s.l. (Hya- dx.doi.org/10.11646​/phyto​taxa.288.2.5
cinthaceae). Phyton (Horn) 38:1–141 Viviani D (1824) Florae libycae specimen. Typographia Pagano, Gen-
Speta F (2001) Die echte und die falsche Meerzwiebel: Charybdis uae [Genova]
Speta und Stellarioides Medicus (Hyacinthaceae), mit Neu- WCSP (2020) World Checklist of Selected Plant Families. Facilitated
beschreibungen und Neukombinationen im Anhang. Stapfia by the Royal Botanic Gardens, Kew. Available at: http://wcsp.
75:139–176 scien​ce.kew.org/. Accessed 10 Jan 2020
Stearn WT (1978) Mediterranean and Indian species of Drimia (Lili- Willdenow CL (1799) Caroli a Linné Species Plantarum. Editio
aceae): a nomenclatural survey with special reference to the Quarta, vol. 2(1). Impensis G.C. Nauk, Berolini [Berlin]
medicinal squill, D. maritima (syn. Urginea maritima). Ann Mus
Goulandris 4:199–210 Publisher’s Note Springer Nature remains neutral with regard to
The Plant List (2020). The Plant List, version 1.1. Available at: http:// jurisdictional claims in published maps and institutional affiliations.
www.thepl​antli​st.org/. Accessed 10 Jan 2020
Thiers B (2020) [continuously updated]. Index Herbariorum: A global
directory of public herbaria and associated staff. New York

13

View publication stats