Marques & al.

• A three-genome phylogeny of Narcissus (Amaryllidaceae) TAXON 66 (4) • August 2017: 832–854

A three-genome five-gene comprehensive phylogeny of the bulbous

genus Narcissus (Amaryllidaceae) challenges current classifications
and reveals multiple hybridization events
Isabel Marques,1,2 Javier Fuertes Aguilar,3 Maria Amélia Martins-Louçao,4 Farideh Moharrek5 &
Gonzalo Nieto Feliner3
1 Department of Agricultural and Environmental Sciences, High Polytechnic School of Huesca, University of Zaragoza, Carretera
Cuarte km 1, 22071 Huesca, Spain
2 UBC Botanical Garden & Centre for Plant Research and Department of Botany, University of British Columbia, 3529-6270
University Blvd, Vancouver BC V6T 1Z4, Canada
3 Real Jardín Botánico, CSIC, Plaza de Murillo 2, Madrid 28014, Spain
4 Centre for Ecology, Evolution and Environmental Changes, Faculdade de Ciências, University of Lisbon. Lisbon, Portugal
5 Department of Plant Biology, Faculty of Biological Sciences, Tarbiat Modares University, Tehran 14115-154, Iran
Author for correspondence: Isabel Marques, isabel.ic@gmail.com
ORCID  IM, http://orcid.org/0000-0001-9788-4831

DOI  https://doi.org/10.12705/664.3

Abstract  Besides being one of the most popular ornamental bulbs in western horticulture, the Mediterranean genus Narcissus
has been the subject of numerous studies focusing on a wide scope of topics, including cytogenetics, hybridization and the
evolution of polymorphic sexual systems. Phylogenetic hypotheses based on chloroplast data have provided a backbone for
the genus but a detailed phylogenetic framework is still lacking. To fill this gap, we present a phylogenetic study of the genus
using five markers from three genomes: ndhF and matK (chloroplast DNA), cob and atpA (mitochondrial DNA), and ITS
(nuclear ribosomal DNA). In addition, we use chromosome counts from 89 populations representing 69 taxa. All analyses
confirm that Narcissus is monophyletic with two main lineages largely corresponding to subg. Hermione and subg. Narcissus,
but with incongruences between organellar and nuclear ribosomal phylogenies. At the infrageneric levels, our phylogenetic
results challenge well-established taxonomic groups, such as sect. Jonquillae, sect. Bulbocodii and sect. Pseudonarcissi, each
of which contains at least two distinct lineages that do not constitute monophyletic groups, and highlight the influence of
allopolyploid species in the monophyly of sections within subg. Hermione. The type of the genus and its section is also nested
within sect. Pseudonarcissi supporting new nomenclatural changes. Our results also confirm the intersubgeneric hybrid nature
of several hybrids including allopolyploids (e.g., N. dubius, N. tortifolius, N. miniatus). Morphological and cytogenetic evidence
independently support the hypothesis that some of the incongruence can be attributed to hybridization, such as the splits of
sect. Bulbocodii and sect. Pseudonarcissi or the disparate phylogenetic placements of sect. Aurelia and sect. Ganymedes.
Together, this indicates a significant role for reticulate evolution in shaping the diversity of this genus. A Bayesian divergence
time analysis suggests that the major diversification events took place during the Neogene and provides younger estimates for
the main nodes than previous studies, which fit paleoclimatic and paleotectonic reconstructions of the western Mediterranean
during this period.

Keywords  hybridization; ITS; Mediterranean region; Narcissus; organellar DNA; phylogeny

Supplementary Material  Electronic Supplement (App. S1–S6; Tables S1–S7; Figs. S1–S9) and DNA sequence alignment files are
available in the Supplementary Data section of the online version of this article at http://ingentaconnect.com/content/iapt/tax

INTRODUCTION important commercial bulbous plants. Moreover, it is one of

Narcissus L. (Amaryllidaceae) is a medium-sized genus of
geophytes mainly distributed in the Mediterranean basin. The
center of diversity lies in the Western Mediterranean area, par-
ticularly in the Iberian Peninsula (Blanchard, 1990; Mathew,
2002). The genus has been reported to have originated in the
Late Oligocene to Early Miocene eras, around 29.3–18.1 Ma
(Santos-Gally & al., 2012) and is currently one of the most
the oldest groups of cultivated bulbs, with more than 27,000
names registered (Hanks, 2002; Kington, 2008).
Following Fernandes’s (1934, 1951) pioneering cytogenetic
research, Narcissus has been a model for karyotype evolution
involving not only chromosome number changes but also sig-
nificant structural chromosomal rearrangements (Fernandes &
Neves, 1941; Fernandes, 1968, 1975; Díaz Lifante & al., 2009)
and genome size changes (Zonneveld, 2008; Marques & al.,

Received: 8 Sep 2016 | returned for (first) revision: 15 Dec 2016 | (last) revision received: 9 May 2017 | accepted: 10 May 2017 || publication date(s):
online fast track, n/a; in print and online issues, 18 Aug 2017 || © International Association for Plant Taxonomy (IAPT) 2017

832 Version of Record

TAXON 66 (4) • August 2017: 832–854
2012a). Since the early reports of heterostyly by Fernandes
(e.g., 1935, 1964), initially questioned by Bateman (1952) and
Baker (1964), the reproductive biology of Narcissus species
has also attracted much attention. In particular, a considerable
number of studies have focused on the evolution of hetero-
styly in this genus (e.g., Arroyo & Dafni, 1995; Barrett & al.,
1996, 2004; Pérez & al., 2004, 2006; Barrett & Harder 2005;
Medrano & al., 2005). Other subjects of plant evolution that
have been addressed using Narcissus species include natural
hybridization (Marques & al., 2007, 2011, 2012b, 2016) as well
as differentiation and other processes at the micro-evolutionary
level (Jiménez & al., 2009; Medrano & al., 2014). In addition
to being a model system for basic research, Narcissus has also
been the subject of drug screening focused on diseases such
as Alzheimer’s (Heinrich & Teoh, 2004; Vrondeli & al., 2005;
Rønsted & al., 2008; Berkov & al., 2014).
However, despite all this progress, the taxonomy of Nar­
cissus remains unsettled due to a number of factors, including
the apparent importance of natural hybridization, the pheno-
typic variability of some morphological traits and the possible
naturalization of some cultivars (Fernandes, 1968; Blanchard,
1990; Mathew, 2002). Historically, both wide and narrow spe-
cies concepts have been proposed and, therefore, estimates of
the number of species in Narcissus have varied extensively
through time, from between 16 to nearly 160 (e.g., Haworth,
1831; Herbert, 1837; Spach, 1846; Baker, 1875). Much of the
disagreement lies in the recognition of closely related taxa
either as different subspecies, varieties or simply as the same
species adapted to different environments, a disparity that is
still visible in modern classifications (Electr. Suppl.: Table S1).
By 1968, Fernandes had accepted 63 species divided into 10
sections (Fernandes, 1968), which were later reduced to 27
species in 11 sections by Webb, who prioritized morphology
over cytological data (Webb, 1980). The subsequent classifica-
tions mostly followed the approach of Fernandes but incorpo-
rated newly described species (e.g., Blanchard, 1990: 65 spp.;
Mathew, 2002: 87 spp.; RHS, 2016: 81 spp.). However, synthetic
approaches continue to be published (Zonneveld, 2008: 36 spp.;
Aedo, 2013: 25 spp.) with discrepancies occurring at the sec-
tional level or even in the recognition or not of more than one
subgenus (Electr. Suppl.: Table S1).
The phylogeny of Narcissus also remains relatively un-
settled. The first small-scale phylogenetic analysis includ-
ing species from all sections based on chloroplast markers,
found only one section to be clearly monophyletic (Graham
& Barrett, 2004). It also supported the division of the genus
into two clades, largely corresponding to N. subg. Hermione
(Haw.) Spach and subg. Narcissus (Graham & Barrett, 2004).
An extended phylogenetic analysis with 39 species, also using
chloroplast markers (Rønsted & al., 2008), confirmed the topol-
ogy of the previous phylogeny. Using nuclear DNA content,
Zonneveld (2008) proposed a new classification eliminating
sections and creating new ones, solving most of the paraphyly
found in those phylogenies. However, the most recent phyloge-
netic work found even subg. Hermione to be non-monophyletic
(Santos-Gally & al., 2012). A more comprehensive phylogenetic
framework based on a wide sampling and using differently
Version of Record
Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae)
inherited markers is thus needed, in particular because hybrid-
ization between Narcissus species has been frequently reported
(Fernandes, 1967; Graham & Barrett 2004; Marques & al.,
2007). For instance, 40% of specific and infraspecific names
of naturally growing plants in the latest list of accepted botani-
cal names of Narcissus have a suspected or explicitly attrib-
uted hybrid origin (https://www.rhs.org.uk/plants/pdfs/plant-
registration-forms/daffalpha.pdf).
Generating a phylogenetic hypothesis that serves as a
sound evolutionary framework in which hypotheses can be
tested, requires that several issues be addressed. A representa-
tive sampling of species, geographic areas, and sequences from
genomic regions with different modes of inheritance should
ideally go along with accounting for the main processes re-
sponsible for the differences between gene trees and species
trees (i.e., hybridization; incomplete lineage sorting [ILS]) and
for those causing mixed phylogenetic signals (recombination).
In addition, exploring divergence times and diversification
rate shifts, together with cytogenetic data and morphological
characters, may help in our understanding of the topological
patterns retrieved from phylogenetic analyses and ultimately in
interpreting the diversity and systematic implications.
Therefore, our general goal was to construct the first multi-
locus phylogeny of Narcissus that samples three genomes, spe-
cifically DNA sequences from the nuclear ribosomal internal
transcribed spacer (ITS) regions (ITS1 + 5.8S + ITS2), two chlo-
roplast regions (matK, ndhF), and two mitochondrial regions
(atpA, cob). To achieve this in this genus, we focus on the follow-
ing issues that critically influence the phylogenetic signal and
its relevance to a natural classification: (1) integrating cytoge-
netic evidence (our own and that available in the literature); (2)
assessing hybridization and its consequences on the systematic
arrangement in the genus; (3) estimating divergence times and
diversification rate shifts; and (4) testing currently recognized
infrageneric taxonomic groups (subgenera, sections).
MATERIALS AND METHODS
Plant sampling for phylogenetic and karyological studies.
— For the phylogenetic study, we collected 1770 specimens of
Narcissus belonging to 103 taxa and 277 populations covering
the whole geographic distribution of Narcissus, the range of
morphological variation and all of the currently recognized
sections (Appendix 1; Electr. Suppl.: Tables S2, S3): N. sect.
Aurelia (Gay) Baker, sect. Apodanthi A.Fern., sect. Braxireon
(Raf.) Valdés (= sect. Tapeinanthus (Herb.) Traub), sect. Bulbo­
codii DC., sect. Ganymedes (Salisb. ex Haw) Schult. & Schult.f.,
sect. Jonquillae DC., sect. Narcissus, sect. Pseudonarcissi DC.,
sect. Serotini Parl. and sect. Tazettae DC. Specimens were
identified based on the taxonomic keys of Fernandes (1968)
except for additional species described later (Electr. Suppl.:
Table S3). Twenty-one putative hybrid taxa were included
in this study: two species suggested to have a hybrid origin
(N. dubius Gouan, N. tortifolius Fern.Casas), fourteen accepted
hybrids (e.g., N. ×rozeirae Pérez-Chisc. & Fern.Casas), plus
six taxa collected in the wild and having clearly intermediate
833
Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae)
morphological characters between two sympatric species (e.g.,
N. fernandesii Pedro × N. papyraceus Ker Gawl.). Based on
their close position to Narcissus, four Amaryllidaceae species
were used as outgroups: Lapiedra martinezii Lag., Leucojum
aestivum L., Nerine bowdenii Watson and Sternbergia lutea
(L.) Ker-Gawl. ex Spreng. (Ito & al., 1999; Meerow & al., 1999;
Lledó & al., 2004; Meerow & Snijman 2006; Meerow & al.,
2006). Samples were preserved as bulbs, silica-gel dried leaves
and/or stored at −80°C until DNA isolation.
Of the 277 populations included in this study, chromosome
numbers were obtained from 89 of them covering 69 of the 103
taxa sampled. For each taxon, root tips were collected from 8 to
15 different individuals (Electr. Suppl.: Table S3), either using
radicles from germinated seeds or directly from bulb roots.
Roots were prepared according to Fernandes (1966, 1975). Five
to ten different metaphase plates per individual were observed
in a total of 2990 observations performed from 2005 to 2010.
Previous reports in the literature were also considered in cases
where either a voucher specimen or a confident geographic
origin could be traced.
DNA isolation, PCR amplification, cloning and sequencing.
— DNA was isolated using the DNeasy Plant Minikit (Qiagen,
Hilden, Germany) following the manufacturer’s instructions.
DNA concentration was measured spectrophotometrically using
a GeneQuant RNA/DNA calculator with a capillary cuvette
(Pharmacia, Cambridge, U.K.) and subsequently preserved at
−80°C. Two chloroplast regions (ndhF, matK), two mitochon-
drial regions (cob, atpA) and one nuclear region (ITS including
ITS1 + 5.8S + ITS2) were amplified using primers and conditions
summarized in Marques & al. (2010). Since direct sequencing
recurrently produced multiple sequence signals, purified PCR
products of the ITS region were all cloned following Marques
& al. (2010) using the One Shot TOP10 protocol (Invitrogen,
San Diego, California, U.S.A.). However, because of the high
number of specimens included in this study and to assure that
all ITS sequences were cloned, we were only able to sample one
individual per population. Competent TOP10F’cells were chem-
ically transformed and the resulting colonies were screened
for plasmids with inserts by PCR, isolating 10 single colonies
from each reaction. Purified PCR products were sequenced in
both directions on a 3730 DNA Analyzer (Applied Biosystems,
Foster City, California, U.S.A.) at the Centro de Genómica y
Proteómica, Parque Científico in Madrid (Spain). In total, we
obtained 1770 sequences (1750 of Narcissus plus 20 sequences
for the outgroups) from each of the four organellar regions repre-
senting all specimens collected for this study. For the ITS region,
we obtained 1640 cloned sequences (1600 from Narcissus and
40 from the outgroups) from 163 individuals (Electr. Suppl.:
Table S3). DnaSP v.5.10.1 (Librado & Rozas, 2009) was used to
characterize DNA polymorphism and nucleotide diversity (π).
Sequence data quality tests. — Checking the reliability
of sequence data is always necessary but particularly so when
using large datasets. Thus, we (1) tested the existence of re-
combinants in all datasets (Schierup & Hein, 2000a, b; Posada
& Crandall, 2002; Electr. Suppl.: Appendix S1), (2) we looked
for possible pseudogenes in the ITS dataset, which if unde-
tected would result in analyzing paralogous sequences (Mayol &
834 Version of Record
TAXON 66 (4) • August 2017: 832–854
Rosselló, 2001; Bailey & al., 2003; Electr. Suppl.: Appendix S2),
and (3) we performed distance-matrix rate tests to graphically
explore the disparities in the evolution of different genes from
the same organisms without previous knowledge of the time of
divergence between species or their phylogenetic relationships
(Syvanen, 2002; Hao & Golding, 2006; Electr. Suppl.: Appendix
S3). We also constructed parsimony networks of the chloroplast
and mitochondrial datasets using TCS v.1.21 (Clement & al.,
2000) for obtaining detailed information and visualization of
the haplotype distribution across species with underlying taxo-
nomic information (Electr. Suppl.: Appendix S4).
Phylogenetic analyses. — Consensus sequences from elec-
tropherograms were assembled and sequences were aligned
manually using BioEdit v.7.0.0 Sequence Alignment Editor
(Hall, 1999). Three matrices were obtained (chloroplast, mito-
chondrial, nuclear) and analyzed to detect variable sites and
informative positions. Since no phylogenetic incongruence
was found between any of the four organellar regions (Electr.
Suppl.: Appendix S5 & Table S4), they were all combined into
a single matrix. In contrast, the topological conflict between
the ITS and the organellar regions involving well-supported
lineages meant that these datasets could not be combined
(Electr. Suppl.: Appendix S5 & Table S4). For efficiency in
the phylogenetic analyses, we minimized redundant sequences,
eliminating identical sequences from the same population in
the organellar dataset. In the case of ITS, in which only one
individual per population was sampled, one copy of each of the
different cloned sequences from each individual was retained;
in practice, this meant one in most cases and two in a few cases.
As a result, from all 1770 organellar sequences, we obtained
a non-redundant matrix containing 289 organellar sequences,
and from all 1650 cloned individuals, we obtained a non-​
redundant matrix with 169 ITS sequences. These sequences
were deposited in GenBank (Appendix 1; Electr. Suppl.: Table
S3). To check for the influence of allopolyploid taxa on the non-
monophyly of clades, N. dubius, N. miniatus and N. tortifolius
were excluded to create an organellar non-redundant matrix
without allopolyploids. For ITS, sequences from the hybrids
N. ×alleniae and N. ×pujolii were also excluded to create a
non-redundant matrix without allopolyploids. In addition, a
smaller matrix containing only chloroplast sequences (the re-
duced chloroplast ndhF-matK dataset) was used for estimat-
ing divergence times and diversification rate shifts. This was
obtained by trimming the non-redundant chloroplast matrix
to include only one sequence per species (or subspecies if any)
and when possible retaining the earliest diverging sequence.
Sequences from non-established hybrids were also excluded
from this reduced matrix that contained 105 sequences, but
two extra outgroup species (Asparagus cochinchinensis (Lour.)
Merr., Nothoscordum dialystemon (Guagl.) Crosa) were added
to calibrate the trees.
Maximum parsimony (MP) analyses were conducted with
PAUP v.4.0a152 for Macintosh (Swofford, 1999) using the
heuristic search option with 100 replicate searches, each with
random addition of sequences and tree bisection-reconnection
(TBR). In each replicate, 50 parsimonious trees were saved
with length equal to or shorter than the most parsimonious
TAXON 66 (4) • August 2017: 832–854 Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae)

trees obtained in a single addition heuristic search. Clade sup-
port was estimated using 104 fast stepwise-addition replicates
(Mort & al., 2000).
The best available model of molecular evolution for each
DNA region was selected using hierarchical likelihood ratio
tests (hLRT) and Akaike information criterion (AIC) as imple-
mented in MrModeltest v.2.3 (Nylander, 2002). GTR + I + G
(Yang, 1996), the symmetrical model with some sites assumed
to be invariable and variable sites assumed to follow a discrete
gamma distribution, was the best fit for all datasets. This model
was used to perform Bayesian analyses using MrBayes v.3.2.6
(Huelsenbeck & Ronquist, 2001; Ronquist & al., 2012). Four
Markov chain Monte Carlo (MCMC) simulations were run
simultaneously for 50 × 106 generations and sampled every 100
generations. We confirmed that the likelihoods had reached
a steady phase with Tracer v.1.3.1 (Rambaut & Drummond,
2007), and 25% of the total number of trees was discarded
as burn-in. The remaining sample was used to create a 50%
majority-rule consensus tree. The program FigTree v.1.4 (http://
tree.bio.ed.ac.uk/software/figtree/) was used for visualization
of trees obtained from parsimony and Bayesian analyses.
Molecular dating. — Absolute lineage divergence times in
Narcissus were estimated in BEAST (Drummond & al., 2012)
using a Bayesian relaxed clock-model. We used the reduced
chloroplast ndhF-matK dataset (N = 105) because the level of
variation of the mitochondrial data is too low and we opted not
to use ITS sequences since generation times and polyploidy
may affect rates of homogenization (Álvarez & Wendel, 2003)
and thus the molecular clock. We used the following settings:
a non-partitioned dataset with GTR + I + G as a substitution
model, birth-death speciation model, and uncorrelated lognor-
mal relaxed clock (UCLD). Two replicate MCMC searches of
30 million generations each were run under these settings and
their results pooled using LogCombiner v.1.7.2 (after removing
25% of samples as burn-in). We used Tracer v.1.6 (Rambaut &
al., 2014) to determine stationarity of the Markov chain and to
verify that all parameters had effective sampling sizes (ESSs)
> 200. TreeAnnotator v.1.4.2 and FigTree v.1.3.1, respectively,
were used to produce and visualize the resulting maximum
clade credibility (MCC) tree. Two external calibration points
based on secondary metacalibration of Angiosperm clades
(Magallón & al., 2015) were used to obtain absolute divergence
times: (a) The root node, i.e., the crown age of Amaryllidaceae
or the divergence between Asparagus L. and the rest of the tree,
was constrained to 62.49 (49.18–76.72) Ma and (b) the stem age
of Amaryllidoideae, as inferred from the divergence time esti-
mated between subfamilies Allioideae and Amaryllidoideae,
represented by Nothoscordum L. and Crinum L., respectively,
in Magallón & al. (2015), was set at 55.21 (40.23–70.17) Ma. In
our dataset, this node is represented by the divergence between
Nothoscordum and all Amaryllidoideae genera: Nerine Herb.,
Lapiedra Lag., Leucojum L., Sternbergia Waldst. & Kit., and
Narcissus.
RESULTS
Sequence data quality tests. — Tests for recombination in
ITS sequences detected significant events only for 13 accessions
(Electr. Suppl.: Appendix S1 & Table S5), ten of them falling in
the same clade in the BI ITS tree. No recombination events were
detected in the chloroplast or mitochondrial sequences. With
respect to the possible occurrence of pseudogenes, inspection
of conserved motifs, secondary models, sequence integrity of
the 5.8S gene, methylation sites (Electr. Suppl.: Fig. S1) and
GC contents (Electr. Suppl.: Fig. S2) revealed that only three
sequences of N. tortifolius 1 could be identified as potential
pseudogenes (Electr. Suppl.: Appendix S2). Since the exclu-
sion of these three sequences in preliminary analyses did not
affect the results, they were maintained in the final matrices.
The mutation rate in ITS differs from that of the organellar
sequences, which was particularly evident when comparing
sequences of subg. Hermione among themselves and against
sequences of subg. Narcissus (Electr. Suppl.: Fig. S3).
Nucleotide characteristics and variation of organellar
sequences. — The aligned organellar DNA matrix was com-
prised of 3285 bp, 1285 from the chloroplast regions and 2000
from the mitochondrial regions (Table 1). Of 178 variable sites,

Table 1. Characteristics of the DNA regions sequenced (outgroups included). Values between brackets in the organellar regions refer to the partial
data for each of the two sequenced regions.
Organellar regions Nuclear regions
Regions cpDNA (ndhF, matK) mitDNA (cob, atpA) Total ITS1 5.8S ITS2 Total
Aligned length [bp] 1285 (738, 547) 2000 (978, 1022) 3285 255 163 247 882
Length range [bp] 1285 (738, 547) 2000 (978, 1022) 3285 230–251 159–163 232–242 841–870
No. of variable characters 144 (97, 47) 34 (10, 24) 178 149 40 132 346
No. of informative characters 73 (43, 30) 10 (9, 1) 83 49 93 40 156
Total number of indels 0 0 0 11 1 15 27
No. of 1-bp indels 0 0 0 3 0 8 11
Maximum length of indel [bp] 0 0 0 12 6 8 –
% of GC content 31.51 (32.55, 30.47) 43.42 (41.21, 45.64) 37.47 70.20 56.41 70.44 63.74
Nucleotide diversity (π) 0.01647 (0.02053, 0.01101) 0.00113 (0.00126, 0.00099) 0.00713 0.10576 0.02114 0.09638 0.06559

Version of Record 835

Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae)
83 were parsimony informative (including outgroups). The
two chloroplast regions contained more informative charac-
ters than the mitochondrial regions (ndhF > matK > cob >
atpA), but the latter were particularly useful for discriminat-
ing sequences between subg. Hermione and subg. Narcissus.
Nucleotide diversity (π) was higher in the chloroplast than
in the mitochondrial regions. No gaps were required for the
alignment of the sequences. Average GC content was higher
in the mitochondrial than in the chloroplast regions (Table 1).
Figure S4 (Electr. Suppl.) illustrates the distribution of plastid
DNA haplotypes across species in a statistical parsimony net-
work constructed using the algorithm described in Templeton
& al. (1992).
Phylogenetic analyses of organellar sequences. — BI
analyses based on 289 sequences generated 24,010 trees for
the combined non-redundant organellar DNA matrix with a
log likelihood (ln L) score of −10,239.23. Values of the poten-
tial scale reduction factor (PSRF) were always equal or close
to 1. MP analysis yielded 283 equally parsimonious trees of
352 steps with a consistency index (CI) of 0.74 excluding non-​
informative characters and a retention Index (RI) of 0.86.
Because BI and MP analyses of the organellar sequences
yielded identical topologies and comparable levels of branch
support, only BI trees are shown.
Narcissus is recognized as a monophyletic lineage with
high support values (100% PP [posterior probability], 98% BS
[bootstrap support]) and two large, highly supported lineages
(Fig. 1). The first lineage (100% in both analyses) largely cor-
responds to subg. Hermione and included taxa with a main
basic chromosome number x = 5 and a likely derived one of x =
11 (Fig. 1). The second lineage corresponds to subg. Narcissus
(100% PP, 100% BS) and contained all the remaining taxa with
the main basic chromosome number x = 7 and a likely derived
one of x = 13 (Fig. 1).
Subgenus Hermione is divided into two highly supported
groups (I and II in Fig. 1A). Group I (100% PP, 78% BS) in-
cludes most species of sect. Tazettae with the exception of
N. elegans (Haw.) Spach and some specimens of N. tazetta L.
Group II (77% PP, 67% BS) includes three highly supported
subgroups: (1) the monospecific sect. Serotini (100% in both
analysis) including N. serotinus L. and the hybrid N. ×alen-
tejanus Fern.Casas (N. cavanillesii × N. serotinus); (2) a het-
erogeneous group (100% in both analyses) containing the re-
mainder of the accessions of N. tazetta L. plus N. elegans, the
allopolyploid N. miniatus Donn.-Morg. & al. (N. elegans ×
N. serotinus), and two hybrids, N. ×alleniae Donnison-Morgan
(N. miniatus × N. viridiflorus) and N. ×perezlarae Font Quer
(N. cavanillesii × N. miniatus); and (3) the monotypic sect.
Aurelia (100% in both analyses) with N. broussonetii Lag.,
which was placed in a trichotomy with the two other groups
in the Bayesian analysis (99% PP), but was resolved as sister
to them in the parsimony analysis (79% BS).
The second main lineage (subg. Narcissus) is divided into
six major groups (III–VIII). Group III (Fig. 1A; 100% in both
analyses) is sister to the other groups (IV–VIII; 100% PP, 82%
BS) and includes only sect. Apodanthi. Groups IV–VIII are
placed in two large clades (IV + V and VI–VIII) (Fig. 1A, B).
836 Version of Record
TAXON 66 (4) • August 2017: 832–854
The first clade (groups IV + V, 93% PP, 78% BS) includes a sin-
gle accession of N. bulbocodium subsp. quintanilhae A.Fern. 2,
isolated from the other accessions of N. bulbocodium L. (100%
in both analyses), which is sister to a clade (100% in both anal-
yses) showing a trichotomy at its base consisting of (1) one
supported clade (100% in both analyses) including N. viridi-
florus Schousb. (sect. Jonquillae) plus one accession of the
hybrid N. ×alleniae (N. miniatus × N. viridiflorus); (2) one
moderately supported clade (82% PP, 67% BS) containing two
subgroups with some samples from sect. Jonquillae (100%
PP, 71% BS) and those of N. dubius and N. tortifolius from
sect. Tazettae (83% PP, 67% BS; note that they are classified
as subg. Hermione) plus the hybrid N. ×pujolii (N. assoanus
× N. dubius). These two groups, (1) and (2), are collectively
labeled as IV (Fig. 1A). (3) The third, well-supported, lineage
(100% PP, 82% BS; labeled as Group V; Fig. 1A) includes most
specimens from sect. Bulbocodii.
The second large clade of subg. Narcissus (groups VI–
VIII; 68% PP, 57% BS; Fig. 1B) includes two weakly supported
clades (VI and VII + VIII). Group VI includes the remainder of
the species of sect. Bulbocodii (76% PP, 54% BS; Fig. 1B). The
clade encompassing groups VII and VIII is weakly supported
(68% PP, 57% BS) and has a trichotomy at its base. Two of the
three branches of that trichotomy are named as group VII, the
first including sect. Braxireon with N. cavanillesii Barra &
G.López and its two hybrids (100% in both analyses), and the
second branch containing two clades, one encompassing the
remainder of the species of sect. Jonquillae (100% PP, 86% BS)
and the other including accessions of sect. Pseudonarcissi from
southern Spain (100% PP, 85% BS; Fig. 1B). Group VIII is a
polytomy (100% in both analyses) composed of the two species
of sect. Narcissus (not forming a clade), the monophyletic sect.
Ganymedes (100% in both analyses) and the remaining taxa of
sect. Pseudonarcissi arranged in two different placements (Fig.
1B). Several subspecies (and varieties) do not fall together and
should be further investigated.
Nucleotide characteristics and variation of ITS sequences.
— Total length of the ITS1 + 5.8S + ITS2 sequences varied from
841 to 870 bp. ITS1 and ITS2 showed a comparable length
and number of variable and informative characters (Table 1).
Alignment of the matrix required the introduction of gaps,
especially in sequences belonging to subg. Hermione. The
largest indels required a 12-bp deletion in ITS1 in six acces-
sions of N. ×alleniae, N. elegans, N. miniatus and, N. serotinus
and an 8-bp insertion occurring in the ITS2 of 39 sequences
of subg. Hermione and sect. Apodanthi. The number of 1-bp
indels was higher in ITS2 than in ITS1 (Table 1) because 8 of
the 11 indels needed for outgroup alignment are concentrated
in that region. The 5.8S region had a total length of 163 bp
except in three sequences of N. tortifolius 1 (Electr. Suppl.:
Appendix S2). Average GC content was 63.74%. Sequences of
subg. Hermione had higher values of nucleotide diversity than
those of subg. Narcissus (0.07433 vs. 0.00554). Only six of the
163 individuals sampled rendered more than one different ITS
sequence, four of them corresponding to hybrids or species of
hybrid origin (N. ×alleniae, N. ×pujolii, N. ×rozeirae, N. torti­
folius). Conversely, the same cloned sequence was found in
TAXON 66 (4) • August 2017: 832–854 Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae)
different taxa in 31 cases; in 16 of them one of the individuals
sharing the sequence was a hybrid (Electr. Suppl.: Table S6).
Despite their shorter length, ITS sequences yielded higher
nucleotide diversity than the organellar sequences (0.066 vs.
0.007).
Phylogenetic analyses of ITS sequences. — Bayesian anal-
yses of 169 ITS cloned sequences produced 17,537 trees with a
ln L score of −12,413.67. PSRF values were always close to 1.
MP analysis yielded 25 equally parsimonious trees of 528 steps
with a consistency index (CI) of 0.85 excluding non-informative
characters and a retention index (RI) of 0.99. Since the topol-
ogy of the BI tree is the same as that of the MP tree, only the
BI tree is shown. As in the organellar tree, two major, strongly
supported lineages were recovered that largely correspond to
subg. Hermione and subg. Narcissus (Fig. 2), although the
composition of the groups indicated in roman numerals differ
between the two trees.
In the subg. Hermione lineage (100% PP, 96% BS), two
major groups were retrieved (labeled as group I and II in Fig.
2A) that were not congruent with those found in the organel-
lar DNA tree. Group I (100% in both analyses) included sect.
Serotini in addition to N. elegans, N. miniatus and the hybrid
N. ×alleniae (Fig. 2A). Group II (100% in both analysis) in-
cluded the remainder of the species of sect. Tazettae and sect.
Aurelia (Figs. 1, 2).
The subg. Narcissus lineage (100% PP, 87% BS) was di-
vided into five major groups (III–VII). Group III (84% PP,
75% BS; Fig. 2A) included all species of sect. Apodanthi and,
as in the organellar tree, was sister to the remainder of the
groups. Group IV (100% PP, 90% BS; Fig. 2A) included some
but not all species of sect. Bulbocodii and was sister (100%
PP, 88% BS) to a large poorly supported clade (group V to
VII; 82% PP, 88% BS) that involved a polytomy at its base in
which most branches correspond to the remaining species of
sect. Bulbocodii (group V). The two remaining branches of
the polytomy corresponded to groups VI and VII (Fig. 2A, B).
Group VI is poorly supported (66% PP, 56% BS) and includes
two lineages: one formed by sect. Braxireon (100% in both
analyses) and by some species of sect. Jonquillae (100% PP,
85% BS), and the other containing the remainder of the species
of sect. Jonquillae together with sect. Ganymedes (100% PP,
98% BS) nested within them. Group VII included all species
of sect. Pseudonarcissi (86% PP, 88% BS; Fig. 2B), unlike the
organellar tree where they were distributed in clades VII and
VIII (Fig. 1B).
Removing potential pseudogenes or recombinant se-
quences did not affect the topology of the ITS tree or the main
results as shown in the two inserts (Fig. 2A, B). Interestingly,
of 163 cloned individuals only 6 contained more than one ITS
copy (see sample names shaded in Fig. 2A, B). Some of these
copies fell into different clades and mostly corresponded to
taxa of hybrid origin.
Incongruence between ITS and organellar trees. — The
comparison of the topologies of the organellar and ITS trees
revealed that several well-supported groups differ in their phy-
logenetic placement (Fig. 3). The most evident discordance is
the placement of N. tortifolius and N. dubius which fell into dif-
Version of Record
ferent subgenera in the two trees. These two species, classically
considered part of sect. Tazettae will be discussed below. Five
other supported groups differed in their placement between
the two phylogenies: sect. Serotini, sect. Ganymedes, sect.
Jonquillae, sect. Bulbocodii and sect. Pseudonarcissi (Fig. 3).
To sort out the incongruence between the ITS and organ-
ellar trees we distinguish between discrepancies in taxon in-
tegrity and in phylogenetic position. The incongruence that is
mainly due to phylogenetic position is arranged into different
categories depending on two criteria. First, the extent of the
incongruence, i.e., whether it affects a single individual or
most samples from one species or a group of species forming
one lineage. Second, the degree of the incongruence estimated
in terms of the disparity, or phylogenetic distance, between
the positions where the element in question (sample, species,
multi-species lineage) falls in the two phylogenies. We refer
to deep incongruence when the positions in the two phylog-
enies are in different major lineages (subg. Hermione or subg.
Narcissus). Medium incongruence refers to a disparate posi-
tion of the element in question in the two phylogenies that
involves different small or medium lineages within the same
major lineage. Our phylogenetic study of Narcissus includes
cases that can be referred to those categories, and are noted
in the discussion section.
Karyological study. — Chromosome counts confirmed two
basic chromosome numbers: x = 5 for subg. Hermione and
x = 7 for subg. Narcissus with several derived numbers (Electr.
Suppl.: Table S3). Within subg. Hermione, diploid, tetraploid
and hexaploid levels were recorded. A derived basic number
x = 11 was found in N. papyraceus Ker Gawl., s.l. (2n = 22) and
N. broussonetii Lag. (2n = 22, 44); the two species have a simi-
lar chromosome structure differing only in two pairs of chro-
mosomes. Within subg. Narcissus, polyploidy was detected
in all sections except sect. Apodanthi and sect. Ganymedes
which are diploid throughout (2n = 14). Nevertheless, diploid
cytotypes (74%) outnumbered polyploid cytotypes (Electr.
Suppl.: Table S3). Up to four different chromosome numbers
were found in a single population of N. bulbocodium, which
shows the highest chromosome number (2n = 72) in this study.
Derived chromosome numbers also occur in several popula-
tions of N. bulbocodium subsp. obesus (Salisb.) Maire (2n = 26,
39). B chromosomes were found in several individuals from dif-
ferent sections (Electr. Suppl.: Table S3). Remarkably, N. poeti-
cus L. and N. radiiflorus Salisb. are diploid with chromosomes
similar to the diploid species of sect. Pseudonarcissi.
Putative hybrids have either the same chromosome number
as their parents (e.g., N. asturiensis (Jord.) Pugsley × N. bulbo­
codium subsp. obesus, 2n = 14) or an intermediate chromosome
number (e.g., N. bulbocodium subsp. obesus × N. calcicola
Mendonça 2n = 26), but allopolyploids have their own chromo-
some numbers: N. dubius (2n = 50), N. tortifolius (2n = 36),
N. miniatus (2n = 30).
Molecular dating. — The Bayesian divergence time analy-
sis performed in BEAST provided time estimates of 16.05 Ma
(10.56–23.51, 95% HPD) for the stem age of Narcissus, dating
the origin of the genus to the Middle Miocene (Electr. Suppl.:
Fig. S5). The age for the split into two subgenera (crown age) is
837
Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae) N. bulbocodium 1
TAXON 66 (4) • August 2017: 832–854
Fig. 1A N. bulbocodium 2
N. bulbocodium 3
N. bulbocodium 4
N. bulbocodium 5
N. bulbocodium 6
N. bulbocodium 7
N. bulbocodium 8
Aurelia N. bulbocodium 9
N. bulbocodium 10
Apodanthi N. bulbocodium 11
Braxireon N. tenuifolius
N. bulbocodium 12
Bulbocodii N. bulbocodium 13
Ganymedes N. bulbocodium 14
Jonquillae
Narcissus
N. bulbocodium subsp. obesus 10
N. bulbocodium subsp. obesus 11
N. bulbocodium var. citrinus V
N. ×felineri (bulbocodium × primigenius)
Pseudonarcissi N. ×brevitubulosus (asturiensis × bulbocodium
Serotini N. bulbocodium var. nivalis 1 var. nivalis)
N. bulbocodium 15
Tazettae N. bulbocodium 16
N. bulbocodium var. nivalis 2
N. jacquemoudii
N. bulbocodium subsp. graellsii
N. bulbocodium 17
N. bulbocodium 18
N. bulbocodium subsp. quintanilhae 1
N. bulbocodium subsp. obesus 12
N. jeanmonodii
N. tingitanus

subg. Narcissus
N. bulbocodium 19
N. bulbocodium 20
N. bulbocodium 21
N. ×barrae (bulbocodium × cantabricus)
N. bulbocodium 22
N. bulbocodium 23
N. bulbocodium 24
N. fernandesii 1
N. willkommii 1
N. jonquilla 1
N. jonquilla 2
N. jonquilla 3
N. jonquilla 4
N. jonquilla 5
N. jonquilla 6
N. jonquilla 7
N. jonquilla 8
N. jonquilla var. henriquesii
N. cordubensis
N. fernandesii 2
N. fernandesii 3
N. fernandesii 4
N. fernandesii × N. papyraceus
N. willkommii 2
N. willkommii 3
N. willkommii 4
N. fernandesii var. major
IV
N. cerrolazae
N. dubius 1
N. dubius 2
N. dubius 3
N. dubius 4
N. tortifolius 1
N. tortifolius 2
N. tortifolius 3
N. tortifolius 4
N. ×pujolii (assoanus × dubius)
N. viridiflorus 1
N. viridiflorus 2
N. viridiflorus 3
N. viridiflorus 4
N. viridiflorus 5
N. viridiflorus 6
N. ×alleniae 1 (miniatus × viridiflorus)
N. bulbocodium subsp. quintanilhae 2
N. calcicola 1
N. calcicola 2
N. calcicola 3
N. scaberulus 1
N. scaberulus 2
N. ×carringtonii (scaberulus × triandrus subsp. pallidulus)
N. rupicola 1
N. rupicola 2
N. rupicola 3
N. rupicola 4
N. rupicola 5
N. rupicola × N. triandrus III
N. cuatrecasasii 1
N. cuatrecasasii 2
N. cuatrecasasii var. segimonensis
N. cuatrecasasii × N. triandrus
N. watieri 1
N. watieri 2
N. atlanticus
N. albimarginatus 1
N. albimarginatus 2
N. marvieri 1
N. marvieri 2 N. ×alleniae 1 (miniatus × viridiflorus)
N. ×alleniae 2 (miniatus × viridiflorus)
N. elegans 1
N. elegans 2
N. elegans 3
N. elegans 4
N. elegans 5
N. elegans 6
N. miniatus 1
N. miniatus 2
N. miniatus 3
N. miniatus 4
N. ×perezlarae (cavanillesii × miniatus) II
subg. Hermione

N. tazetta 1
N. tazetta 2
N. tazetta 3
N. broussonetii 1
N. broussonetii 2
N. broussonetii 3
N. broussonetii 1
N. broussonetii 2
N. broussonetii 3
N. ×alentejanus (cavanillesii × serotinus)
N. serotinus 1
N. serotinus 2
N. serotinus 3
N. serotinus 4
N. pannizianus 1
N. pannizianus 2
N. papyraceus 2
N. papyraceus 3
N. papyraceus 4
N. papyraceus 5
N. papyraceus 6
N. polyanthos
N. papyraceus 7 I
N. papyraceus 8
N. pachybolbus
N. papyraceus 9
N. papyraceus 10
N. papyraceus 1
N. tazetta 4
Sternbergia lutea N. tazetta 5
Leucojum aestivum
Lapiedra martinezii
Nerine bowdenii

838 0.0030 Version of Record

TAXON
Fig. 66&(4)
1A (left) • August
B (right). 2017: 832–854
Bayesian tree (50% al. • A three-genome
MarquesN.&asturiensis 1 phylogeny of Narcissus (Amaryllidaceae)
N. asturiensis 2
majority-rule consensus) of organellar N. pseudonarcissus 8
N. asturiensis × N. bulbocodium subsp. obesus
DNA sequences (ndhF and matK chloro- N. asturiensis 3
N. asturiensis 4
plast regions, cob and atpA mitochondrial N. asturiensis 5
N. asturiensis 6
regions) from a non-redundant matrix N. asturiensis 7
N. primigenius 1
containing 289 sequences of Narcissus N. primigenius 2
N. nobilis var. leonensis 1
and 4 outgroups (see text). Posterior N. portensis
N. ×somedanus (asturiensis × triandrus)
probabilities (expressed as percentages) N. tortuosus
N. nobilis var. leonensis 2
and parsimony bootstrap support higher N. cyclamineus 1
N. cyclamineus 2
N. minor 1
than 50% are indicated above and below N. nanus
N. asturiensis 8
branches, respectively. A thicker line N. cyclamineus 3
N. pseudonarcissus 1
indicates 100% support in both analyses. N. pseudonarcissus 2
N. pseudonarcissus 3
Black circles: > 95% support value; grey N. pseudonarcissus 4
N. pseudonarcissus 5
circles: 95 ≥ support value ≥ 70; white N. abscissus 1
N. abscissus 2
circles: < 70 support value. Color-coded N. hispanicus
N. hispanicus subsp. perez-chiscanoi
branches and vertical tags follow cur- N. pseudonarcissus subsp. munozii-garmendiae
N. ×hannibalis (hispanicus × triandrus subsp. lusitanicus)
rently accepted sections as indicated N. nobilis 1
N. nobilis 2
in the legend. Roman numerals denote N. pallidiflorus 1
N. pallidiflorus 2
groups discussed in the main text. N. pseudonarcissus subsp. eugeniae 1
N. pseudonarcissus subsp. eugeniae 2
N. bicolor 1
N. bicolor 2
VIII
N. bujei 1
N. bujei 2
N. jacetanus
N. bujei 3
N. pseudonarcissus subsp. eugeniae 3
N. confusus
N. bujei 4
N. macrolobus
N. pumilus N. triandrus subsp. pallidulus 1
Fig. 1B N. triandrus subsp. pallidulus 2
N. triandrus subsp. pallidulus 3
N. triandrus subsp. pallidulus 4
N. triandrus subsp. pallidulus 5
N. triandrus subsp. pallidulus 6
N. triandrus 1
N. triandrus 2
Aurelia N. triandrus 3
Apodanthi N. triandrus 7
N. triandrus var. loiseleurii
Braxireon N. triandrus 4
Bulbocodii N. ×rozeirae (bulbocodium × triandrus subsp. pallidulus)
N. ×susannae (cantabricus × triandrus subsp. pallidulus)
Ganymedes N. triandrus subsp. lusitanicus 1
Jonquillae N. triandrus subsp. lusitanicus 2

Subg. Narcissus
Narcissus N. triandrus subsp. lusitanicus 3
N. triandrus subsp. lusitanicus 4
Pseudonarcissi N. triandrus × N. bulbocodium
Serotini N. triandrus 5
N. triandrus 6
Tazettae N. bujei 1
N. bujei 2
N. bujei 3
N. bujei 4
N. alpestris
N. moleroi
N. radiiflorus
N. poeticus 1
N. poeticus 2
N. nevadensis 1
N. nevadensis 2
N. nevadensis 3
N. nevadensis 4
N. nevadensis 5
N. nevadensis 6
N. nevadensis 7
N. alcaracensis 1
N. alcaracensis 2
N. alcaracensis 3
N. longispathus 1
N. longispathus 2
N. segurensis
N. yepesii 1
N. longispathus 3
N. longispathus 4
N. longispathus 5
N. yepesii 2
N. yepesii 3
VII
N. yepesii 4
N. assoanus subsp. rivas-martinezii
N. assoanus subsp. praelongus
N. gaditanus 1
N. gaditanus 2
N. gaditanus 3
N. gaditanus 4
N. assoanus 1
N. assoanus 2
N. assoanus 3
N. assoanus 4
N. baeticus
N. cavanillesii 1
N. cavanillesii 2
N. cavanillesii 3
N. cavanillesii 4
N. cavanillesii 5
N. cavanillesii 6
N. ×alentejanus (cavanillesii × serotinus)
N. ×perezlarae (cavanillesii × miniatus)
N. romieuxii 1
N. romieuxii 2
N. romieuxii 1
N. peroccidentalis
N. cantabricus 1
N. cantabricus 2
N. cantabricus 3
N. cantabricus 4
N. hedraeanthus 1
N. hedraeanthus 2
N. ×cazorlanus (hedraeanthus × triandrus subsp. pallidulus)
N. ×tuckeri 1 (fernandesii × hedraeanthus)
N. ×tuckeri 2 (fernandesii × hedraeanthus) VI
N. bulbocodium subsp. obesus 1
N. bulbocodium subsp. obesus 2
N. bulbocodium subsp. obesus 3
N. bulbocodium subsp. obesus 4
N. bulbocodium subsp. obesus 5
N. bulbocodium subsp. obesus × N. calcicola
N. bulbocodium subsp. obesus 6
N. bulbocodium subsp. obesus 7
N. bulbocodium subsp. obesus 8
Version ofN.Record
bulbocodium subsp. obesus 9 839
Marques & al. • A three-genome phylogeny of Narcissus
N. (Amaryllidaceae)
triandrus subsp. pallidulus 1 TAXON 66 (4) • August 2017: 832–854
Fig. 2A N. triandrus subsp. pallidulus 2
N. triandrus 1
N. triandrus subsp. lusitanicus 1
N. triandrus subsp. lusitanicus 2
Aurelia N. triandrus subsp. lusitanicus 3
Apodanthi N. triandrus × bulbocodium N. triandrus 2
N. triandrus 2 N. triandrus 3
Braxireon N. triandrus 3 N. triandrus subsp. loiseleurii
Bulbocodii N. triandrus var. loiseleurii
Ganymedes N. ×rozeirae (bulbocodium × triandrus subsp. pallidulus) rec
Jonquillae N. viridiflorus 1
N. viridiflorus 2
Narcissus N. viridiflorus 3
Pseudonarcissi N. viridiflorus 4
Serotini N. viridiflorus 5
Tazettae N. ×alleniae 1 (miniatus × viridiflorus)
N. jonquilla 1
N. jonquilla 2
/
N. cerrolazae
N. jonquilla var. henriquesii
VI
N. willkommii 1

subg. Narcissus
/ N. fernandesii var. major
N. fernandesii 1
N. fernandesii 2
N. fernandesii × papyraceus
N. cordubensis
N. baeticus
/
N. gaditanus 1
N. gaditanus 2
N. calcicola 1
N. assoanus subsp. praelongus
N. assoanus subsp. rivas-martinezii
N. ×pujolii (assoanus × dubius)
N. assoanus 1
N. assoanus 2
N. cavanillesii 1
N. cavanillesii 2
N. cavanillesii 3
N. cavanillesii 4
N. cavanillesii 5
N. ×alentejanus (N. cavanillesii × N. serotinus)
N. ×perezlarae (N. cavanillesii × N. miniatus)
N. hedraeanthus
N. ×tuckeri (fernandesii × hedraeanthus) rec
N. tenuifolius N. hedraeanthus
N. peroccidentalis N. tenuifolius
N. romieuxii 1 N. peroccidentalis
N. bulbocodium subsp. obesus 1
N. bulbocodium var. citrinus rec V N. romieuxii 1
N. bulbocodium subsp. obesus 1
N. ×barrae (bulbocodium × cantabricus) rec
N. ×cazorlanus (hedraeanthus × triandrus subsp. pallidulus) N. ×cazorlanus (hedraeanthus × triandrus subsp. pallidulus)
N. cantabricus 1
N. cantabricus 1 N. ×susannae (cantabricus × triandrus subsp. pallidulus)
N. ×susannae (cantabricus × triandrus subsp. pallidulus) N. cantabricus 2
N. cantabricus 2
N. bulbocodium subsp. quintanilhae 1 rec
N. bulbocodium subsp. quintanilhae 2 rec
N. bulbocodium 1
N. bulbocodium subsp. graellsii rec
N. jacquemoudii rec
N. tingitanus rec
N. bulbocodium var. nivalis 1 rec
/ N. ×felineri (bulbocodium × primigenius) rec
IV
/
N. ×brevitubulosus (asturiensis × bulbocodium var. nivalis) rec
N. jeanmonodii rec N. bulbocodium 1
N. bulbocodium 2 N. bulbocodium 2
N. bulbocodium subsp. obesus 2 N. bulbocodium subsp. obesus 2
N. bulbocodium subsp. obesus × N. calcicola N. bulbocodium subsp. obesus × N. calcicola
N. ×rozeirae (bulbocodium × triandrus subsp. pallidulus) rec N. bulbocodium subsp. obesus 1
N. bulbocodium subsp. obesus 1
N. cuatrecasasii 1
N. cuatrecasasii var. segimonensis
N. cuatrecasasii 2
N. cuatrecasasii × triandrus
N. atlanticus
N. albimarginatus 1
N. albimarginatus 2
N. rupicola 1
N. rupicola 2 III
N. rupicola × N. triandrus
N. watieri
N. marvieri
/ N. scaberulus 1
/ / N. ×carringtonii (scaberulus × triandrus subsp. lusitanicus)
N. scaberulus 2
N. calcicola 2
N. calcicola 3
/
N. tazetta 1
N. papyraceus 1
N. papyraceus 2
N. pachybolbus
N. polyanthos
/ N. broussonetii 1
N. broussonetii 2
N. papyraceus 3
II
N. papyraceus 4 N. ×pujolii (assoanus × dubius)
/ N. dubius 1
N. pannizianus
subg. Hermione

N. dubius 2
N. ×pujolii (assoanus × dubius) N. tortifolius 2
N. dubius 1
N. dubius 2
N. tortifolius 1*
N. tortifolius 1*
N. tortifolius 2
N. tazetta 2
N. tazetta 3
N. tazetta 4
N. tazetta 5
N. serotinus 1
N. serotinus 2
N. serotinus 3
N. serotinus 4
N. miniatus 1
N. ×alleniae 1 (miniatus × viridiflorus)
N. miniatus 2 I
N. miniatus 3
N. miniatus 4
N. elegans 1
N. elegans 2
N. elegans 3
N. elegans 4
Lapiedra martinezii
Leucojum aestivum
Sternbergia lutea
Nerine bowdenii
840 0. 03 Version of Record
TAXON 66 (4) • August 2017: 832–854 Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae)

13.80 Ma (8.97–19.08, 95% HPD; node 3 in Electr. Suppl.: Fig. DISCUSSION

S5) whereas the initial diversification within subg. Narcissus
and subg. Hermione, i.e., their crown ages, is 9.99 Ma (6.54–
14.12, 95% HPD; node 5) and 5.98 Ma (3.31–9.24, 95% HPD;
node 4), respectively (Electr. Suppl.: Fig. S5). These estimates
for early diversification in Narcissus lie between the Middle
and Late Miocene. Within subg. Narcissus, the majority of
the most diversified lineages, including both groups of sect.
Jonquillae (Electr. Suppl.: Fig. S5, nested within node 12, and
node 7), both groups of sect. Pseudonarcissi (Electr. Suppl.:
Fig. S5, nested within node 10 and 12), and both groups of
sect. Bulbocodii (Electr. Suppl.: Fig. S5, nodes 13 and 14),
diverged in the Pliocene, whereas the less diversified lineages
(sect. Apodanthi and sect. Ganymedes, Electr. Suppl.: Fig. S5,
nodes 6 and 11) have their crown age in the Pleistocene or late
Pliocene. In subg. Hermione, the diversification of its two main
groups occurs at two different times. The crown age for the first
lineage (Electr. Suppl.: Fig. S5, node 9), which includes sect.
Aurelia and sect. Serotini and some species of sect. Tazettae,
can be dated to the Late Pliocene, while the age inferred for
the second group formed exclusively by the remainder of the
species of sect. Tazettae is Early Pleistocene (Electr. Suppl.:
Fig. S5, node 8). We did not detect any significant diversifica-
tion rate-shift in the plastid data (Electr. Suppl.: Fig. S6, Table
S7 & Appendix S6).
Two consistent lineages within Narcissus. — Our phylo-
genetic analyses, based on a comprehensive sampling of se-
quences from the three genomes, confirms the monophyly
of Narcissus, anticipated in Meerow & al. (2006), and the
existence of two lineages largely corresponding to subgenera
Hermione and Narcissus (Fig. 1, 2), also reported in recent
works (Graham & Barrett, 2004; Pérez & al., 2004; Rønsted
& al., 2008; Santos-Gally & al., 2012). These two lineages
were also suggested in Fernandes’s (1968) classical cytogenetic
studies revealing that Hermione has a main basic chromosome
number of x = 5 along with chromosomes of heterogeneous
length, including acrocentric chromosomes. Narcissus has a
main basic chromosome number of x = 7 along with a karyotype
generally composed of long chromosomes and no acrocentric
chromosomes. Apart from the allopolyploids N. dubius and
N. tortifolius (and other established hybrids), which show more
complex patterns due to hybridization (see below), all species
consistently fell either in the Hermione or the Narcissus lineage
in both the nuclear and organellar phylogenies,.
There are also differential morphological, phenological
and distributional patterns. In general, species of the Hermione
lineage flower in autumn or late winter and have tubular flow-
ers with white tepals and a short corona, while representatives

Fig. 2B N. tortuosus
N. pseudonarcissus subsp. eugeniae
N. primigenius
N. nobilis var. leonensis 1
Aurelia N. nobilis 1
N. pumilus
Apodanthi N. pallidiflorus 1
Braxireon N. minor
Bulbocodii N. nanus
N. jacetanus
Ganymedes / N. asturiensis 1
Jonquillae N. ×somedanus (asturiensis × triandrus)
Narcissus N. asturiensis × N. bulbocodium subsp. obesus
N. asturiensis 2
Pseudonarcissi N. radiiflorus
Serotini N. poeticus

subg. Narcissus
Tazettae / N. cyclamineus 1
N. segurensis VII
N. yepesii 1
N. yepesii 2
N. longispathus 1
N. alcaracensis
N. segurensis
N. longispathus 2
/ N. longispathus 3
N. bujei 1
N. bujei 2
N. nevadensis 1
N. nevadensis 2
/ N. nevadensis 3
N. nevadensis 4
N. nevadensis 5
N. pseudonarcissus subsp. perez-chiscanoi
N. pseudonarcissus 1
N. pseudonarcissus subsp. munozii-garmendiae
N. pseudonarcissus 2
N. hispanicus
N. ×hannibalis (hispanicus × triandrus subsp. lusitanicus)
N. alpestris
N. moleroi
N. abscissus
N. macrolobus
N. portensis
N. bicolor 1
N. confusus

Fig. 2A (left) & B (right). Bayesian tree (50% majority-rule consensus) of nuclear ribosomal ITS from a non-redundant matrix containing 169
cloned sequences from 163 samples of Narcissus and 4 outgroups (see text). Posterior probabilities (expressed as percentages) and parsimony
bootstrap support higher than 50% are indicated above and below branches, respectively. A thicker line indicates 100% support in both analyses.
Black circles: > 95% support value; grey circles: 95 ≥ support value ≥ 70; white circles: < 70 support value. Color-coded branches and vertical
tags follow currently accepted sections as indicated in the legend. Roman numerals denote groups discussed in the main text. Shaded names
in blue highlight the positions of different ITS copies from the same cloned individual. Asterisk (*) indicates potential pseudogene sequences;
rec indicates potential recombinant sequences; grey inserts show differences in the ITS tree when potential pseudogenes and recombinant
sequences were removed.

Version of Record 841

Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae) TAXON 66 (4) • August 2017: 832–854

of the subg. Narcissus lineage are mostly spring-flowering
species with cup-shaped or tubular flowers, yellow tepals and a
medium to large corona (Fernandes, 1968). Subgenus Narcissus
is both morphologically more diverse and more species-rich
than subg. Hermione. In addition, species of subg. Hermione
frequently span large ranges extending from the Atlantic coast
of Morocco throughout the Mediterranean basin. Subgenus
Narcissus has diversified in the western Mediterranean and
many species are restricted to the Iberian or Moroccan moun-
tains (Santos-Gally & al., 2012). Thus, based on all this evi-
dence, we recommend classification of Narcissus into the two
subgenera Hermione and Narcissus.
Major systematic conclusions from phylogenetic analyses.
— At the infrageneric level, there is good agreement between
taxonomic assignment and phylogenetic lineages in some
groups (Figs. 1–3). Besides the two subgenera, sect. Apodanthi,
sect. Serotini as well as the monotypic sect. Braxireon appear
as well-defined lineages each with a single origin. The case of
sect. Ganymedes is worth noting since, although it is mono-
phyletic in both the organellar and ITS trees, unexpectedly
its position in the two trees is not the same. It is close to sect.
Narcissus and sect. Pseudonarcissi in a polytomy at the base of
clade VIII in the organellar tree (Fig. 1B), and is nested within
sect. Jonquillae in the ITS tree (Fig. 2A). A similar situation

Pseudonarcissi

orgDNA nrDNA
Pseudonarcissi
Narcissus
VII

Ganymedes
VIII
Pseudonarcissi
Narcissus

Pseudonarcissi
Ganymedes
VII
Jonquillae Jonquillae
Braxireon VI
/

VI Jonquillae
Bulbocodii
Braxireon

V
V Bulbocodii

Bulbocodii
IV
Bulbocodii
IV
Jonquillae

III
Tazettae Apodanthi
Jonquillae
Bulbocodii
III
Apodanthi
Aurelia
II

Tazettae Tazettae
II
outgroups Aurelia outgroups
Serotini Serotini I
Tazettae Tazettae
I

Fig. 3. Major positional and compositional incongruence between 50% majority-rule consensus BI trees from organellar DNA (left) and nrDNA
ITS (right) datasets. Major clades (identified with roman numerals following Figs. 1 and 2) are collapsed, dotted lines link clades (or rarely grades)
of congruent composition in the two phylogenetic trees, and thick solid lines indicate sections which fall into separate clades in one or both
trees. Support for branches follows Figs. 1 and 2. Sections are labeled according to the infrageneric classification we followed at the beginning
of this study (largely based on Fernandes, 1968). See Electr. Suppl.: Fig. S9 for a redrawn figure with the new sectional taxonomy here proposed.

842 Version of Record

TAXON 66 (4) • August 2017: 832–854 Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae)
occurs in sect. Aurelia, which is sister to sect. Serotini and sect.
Tazettae in the organellar tree (Fig. 1A) but is nested within
sect. Tazettae in the ITS tree (Fig. 2A).
In contrast, our phylogenetic analyses identified unex-
pected lineages that question the integrity of four important sec-
tions (sect. Bulbocodii, sect. Jonquillae, sect. Pseudonarcissi,
sect. Tazettae) and defy the current morphologically based
classifications. Two large sections, sect. Jonquillae and sect.
Bulbocodii, consist of two different lineages that are not closest
relative to each other, both in the organellar and nuclear trees;
a third, sect. Pseudonarcissi, is only monophyletic in the ITS
tree; and sect. Tazettae is split into three lineages in the organel-
lar and into two in the ITS trees (Figs. 1–3).
The two clades of sect. Jonquillae are consistent – except
for one uncertain placement (see below) – with the previous
cytogenetic identification of two different types of karyo-
types that were used to define subsect. Jonquillae and subsect.
Juncifolii (Fernandes, 1939). The former grouped karyotypes
with long chromosomes including N. fernandesii, N. jon-
quilla L., N. viridiflorus, and N. willkommii (Samp.) A.Fern.
among others, while the latter grouped karyotypes with short
chromosomes including N. assoanus s.l., N. baeticus Fern.
Casas and N. gaditanus Boiss. & Reut. The only uncertainty in
the matching between the two subsections and our phylogenetic
results is the placement of the only sample of N. baeticus within
subsect. Jonquillae instead of subsect. Juncifolii in the ITS
tree (Fig. 2A). Overall, our results support Fernandes’s groups
(Fernandes, 1939) and question the integrity of sect. Jonquillae.
Section Bulbocodii, encompassing the hoop petticoat daf-
fodils, also consists of two main groups in the two phylogenies
(Figs. 1–3). There are some indications of a geographical pattern
for these two groups in the organellar tree. The group labeled
as V in Fig. 1A includes yellow-flowered species widespread
in the Iberian Peninsula and northern Morocco, with several
ploidy levels. The second group of sect. Bulbocodii, labeled
as VI in Fig. 1B, includes southern Iberian and Moroccan,
mostly white or creamy-yellow-flowered species endemic to
mountains regions. It is likely that these two groups reflect
two previously undetected differentiated lineages of hoop
petticoat daffodils, but an alternative explanation involving
hybridization cannot be totally rejected. Therefore, despite the
clear identification of two lineages, more taxonomic work is
needed to fully understand the diversity, and particularly spe-
cies delimitation, within sect. Bulbocodii.
Section Pseudonarcissi, encompassing the trumpet daffo-
dils, is the third important group of species in which an unex-
pected result was obtained, calling into question its taxonomic
cohesiveness. The pattern and interpretation are less clear than
for sect. Bulbocodii and sect. Jonquillae because in the ITS
phylogeny, all trumpet daffodils fall together within the same
lineage whereas in the organellar phylogeny the pattern is more
complex. In the latter tree, a group of species from southern
Spain forms a clade sister to a part of sect. Jonquillae (Fig. 1B,
group VII) and separate from the northern trumpets. This phy-
logenetic split between southern and northern species of sect.
Pseudonarcissi is also supported by the fact that the two groups
have distinct environmental niches (Marques & al., 2008). The
Version of Record
exception among the southern species in the organellar tree is
N. bujei Fern.Casas of which the sequences fall into two dif-
ferent clades. Some of the sequences are part of a polytomy at
the base of a clade with most northern trumpet daffodils while
the rest are sister to N. alpestris Pugsley in another subclade
within clade VIII (Fig. 1B). An old divergence between the
northern and southern groups and ILS could explain this pat-
tern in which N. bujei would have originated at more or less the
same time when the northern and southern lineages diverged.
However, this hypothesis is compromised by the fact that ILS
is less likely to occur in organellar trees than in nuclear trees
due to lower effective population sizes. Alternatively, a hybrid-
ization event between the southern trumpet lineage and sect.
Jonquillae after the divergence of N. bujei, with subsequent
chloroplast capture, is also a plausible scenario for explaining
the data. Based on genome size data, Zonneveld (2008, 2010)
proposed a new sect. Nevadensis to include N. alcaracensis
Ríos & al., N. bujei, N. longispathus Pugsley, N. nevadensis
Pugsley, N. segurensis and N. yepesii Ríos & al. and also sug-
gested an ancient hybrid origin for N. bujei with N. longispa-
thus and N. hispanicus Gouan as parents. However, the molecu-
lar evidence provided here for N. bujei, which does not easily fit
Zonneveld’s taxonomic proposal (Zonneveld, 2008, 2010), and
the inconclusive morphological evidence recommend further
studies to elucidate the origin of N. bujei (Zonneveld, 2010).
For sect. Tazettae, the organellar–nuclear discordance re-
lated to N. dubius and N. tortifolius is easy to interpret and to
accommodate taxonomically since these two species are inter-
subgeneric allopolyploids (see below). In contrast, the other
distinct lineages of sect. Tazettae do not allow a straightforward
interpretation. The fact that subg. Hermione contains several
allopolyploid species makes it difficult to obtain monophy-
letic lineages unless excluding those allopolyploids. To assess
how allopolyploids could influence the monophyly of sections
in subg. Hermione, we carried out parsimony and Bayesian
analyses of the organellar and nrDNA ITS matrices exclud-
ing N. dubius, N. miniatus and N. tortifolius. In the case of
ITS, we also excluded the hybrids N. ×alleniae and N. ×pujolii
(Figs. 1, 2; Electr. Suppl.: Figs. S7, S8). The results show that
three sections (sect. Serotini, sect. Tazettae, and the monotypic
sect. Angustifolii Fernandes containing N. elegans; see below)
constitute monophyletic groups provided that we disregard the
likely hybridogenous sect. Aurelia (see below).
Our study also has implications for the composition of a
fourth section, sect. Narcissus, which is nested within sect.
Pseudonarcissi which questions its taxonomic integrity; in the
organellar tree the samples of sect. Narcissus do not even form
a monophyletic group. From a karyological and molecular point
of view it is not possible to separate sect. Narcissus from sect.
Pseudonarcissi, but morphologically they are quite distinct.
Species of sect. Narcissus usually have single flowers, pure
white tepals with a discoidal to cup-shaped, intensely red or
orange corona (Blanchard, 1990), which is similar to the floral
morphology of members of subg. Hermione, although no molec-
ular evidence supports this relationship. Other similarities be-
tween N. poeticus and species of sect. Pseudonarcissi are found
in the karyotypes (Fernandes, 1968, 1975) and genome size
843
Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae)
values (Zonneveld, 2010), which are often identical. Therefore,
the evidence available indicates merging of these two sections.
Yet, the fact that N. poeticus, along with the other members of
this section, has a long history of cultivation and artificial selec-
tion, requires caution when discussing relationships.
The fact that several subspecies (and varieties) often do not
group together (e.g., N. bulbocodium subsp. obesus; N. nobi-
lis var. leonensis (Pugsley) A.Fern., among others) should be
further investigated because this may reflect poor taxonomic
placement.
Evidence of hybridization in the evolution of Narcissus. —
In addition to questioning the current circumscription of the four
sections, we found support for the existence of hybridization,
including allopolyploid speciation. Incongruence due to discor-
dant placement of the same taxa in the nuclear and organellar
trees is a significant finding, which has systematic and evo-
lutionary implications. Incongruence between tree topologies
from genes with different inheritance can be due mainly to
hybridization or ILS. Telling apart these two phenomena is com-
plex both for recent and old divergence and hybridization events.
A number of useful approaches have been proposed to tackle
this (Maureira-Butler & al., 2008; Joly & al., 2009; Kubatko
& al., 2009; Pelser & al., 2010; Escudero & al., 2014; Payseur
& Rieseberg, 2016). However, these approaches tend to gener-
ate an excess of false positives, identifying lineages as hybrids
when they are not (Heled & al., 2013). Additionally, they have
mostly been tested on simulated data and/or only on datasets
of very limited size (Jones & al., 2014). Because of the size and
complexity of our datasets, we analyzed the phylogenetic results
together with our own independent evidence based on cyto-
genetic data, morphology and reproductive biology (Marques
& al., 2010, 2012b, 2016), as well as with previous evidence
from various sources (see introduction) in order to identify the
causes of the underlying incongruence (Wendel & Doyle, 1998).
Below we only present highly likely cases of hybridization in
Narcissus revealed by phylogenetic incongruence and supported
by independent data, but further specific studies are needed to
support other potential “minor” cases of hybridization (e.g.,
N. bulbocodium subsp. obesus × N. calcicola, N. ×hannibalis,
N. ×tuckeri, among others) and the potential hybrid origin of
several species (e.g., N. fernandesii). To facilitate discussion,
we group them into deep and medium incongruences following
the categories described in the results. A general observation
resulting from this work is that putative cases of hybridization
are not restricted to closely related species.
Deep incongruence. — Narcissus dubius (2n = 50; distrib-
uted from southeastern Spain to southern France) and N. torti­
folius (2n = 36; endemic to a few populations in southeastern
Spain; Barra & López-González, 1982; Barra, 1999; Sánchez
Gómez & al., 2000) have classically been assigned to sect.
Tazettae based on morphological characters (e.g., Fernandes,
1969, 1975; Webb, 1980), or recognized as a separate sect.
Dubii (Fernández Casas, 1983, 1984). However, in our study
only the ITS tree supports the relationships with sect. Tazettae
(subg. Hermione, Fig. 2A, group II), whereas in the organel-
lar tree, they are placed with other species of sect. Jonquillae
(subg. Narcissus; Fig. 1A, group IV). This pattern supports the
844 Version of Record
TAXON 66 (4) • August 2017: 832–854
hypothesis that both species are intersubgeneric allopolyploids.
Furthermore, our results suggest that plants of sect. Jonquillae
(either an ancestral tetraploid or an unreduced form of N. jon-
quilla) might have acted as female parent and N. papyraceus
(or a close species) as male parent. This parentage is recipro-
cal to that hypothesized earlier based on the close karyologi-
cal and morphological similarity with N. tazetta (Fernandes,
1937, 1969, 1975; Koopowitz & al., 2002; Romero & al., 1983).
The fact that these two allopolyploids are completely fertile
(Koopowitz & al., 2002) suggests that their origin is not recent.
Narcissus ×alentejanus (N. cavanillesii × N. serotinus),
N. ×alleniae (N. miniatus × N. viridiflorus) and N. ×perezlarae
(N. cavanillesii × N. miniatus) — Hybridization, supported
by morphological and experimental independent evidence
explains these three examples of deep incongruence. These
hybrids have organellar sequences identical or very similar to
their female parent (Fig. 1A), and in the ITS phylogeny they
also appear with one or, in the case of N. ×alleniae, two progen-
itors (Fig. 2A). Preservation of both the maternal and paternal
ITS sequences is probably facilitated by the different genomic
location of ribosomal loci in the parental genomes, which be-
long to different subgenera, that can hamper homogenization
of ITS sequences (Nieto Feliner & Rosselló, 2012), although
stabilized hybrids such as N. ×perezlarae can have distinct
ITS copies (Marques & al., 2010). Our chromosome counts
support the hybrid nature in these cases since N. ×alleniae,
N. ×alentejanus and N. ×perezlarae have chromosome num-
bers intermediate between their parents (Electr. Suppl.: Table
S3; see also Marques & al., 2010 for the latter two examples).
Another example of an intersubgeneric hybrid confirmed
by our phylogenetic study is N. ×pujolii originating from
hybridization between N. assoanus and N. dubius. The posi-
tions where the cloned sequences of this hybrid fall in the ITS
phylogeny reflect the placement of its parental species, i.e., with
N. dubius in the subg. Hermione lineage and with N. assoanus
in the subg. Narcissus lineage (Fig. 2A, group II), whereas in
the organellar phylogeny it groups with N. dubius and N. torti­
folius (Fig. 1A, group IV).
Medium incongruence. — In the ITS tree, sect. Ganymedes
groups with sect. Jonquillae, while in the organellar tree it
groups with several trumpet daffodils (sect. Pseudonarcissi)
(Figs. 1B, 2A, 3). An ancient hybridization event to explain
this incongruence is plausible for morphological reasons. The
nodding flowers and reflexed tepals of N. triandrus L. and its
allies resemble some trumpet daffodils, such as N. cyclamineus,
and suggest a member of sect. Pseudonarcissi as one progenitor
of sect. Ganymedes. The close position in the ITS tree together
with its karyotype suggests that the maternal progenitor was
N. jonquilla s.l.
Narcissus miniatus falls into different lineages within
subg. Hermione in the organellar and nuclear phylogenies (Fig.
1A, group II; Fig. 2A, group I) because of its allohexaploid
nature (2n = 30), its progenitors being N. elegans (2n = 20) and
N. serotinus (2n = 10) (Díaz Lifante & al., 2009; Marques &
al., 2010). Despite the earlier interpretation that this species is
a synonym of N. serotinus (Fernandes, 1951, 1968; Phytos &
Kamari, 1974), recent cytogenetic evidence (Díaz Lifante &
TAXON 66 (4) • August 2017: 832–854 Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae)
al., 2009) is consistent with the phylogenetic results provided
here (Figs. 1, 2) and in Marques & al., (2010). In the ITS tree,
sequences of N. elegans, N. miniatus and N. serotinus fall to-
gether in the same branch, while in the organellar phylogeny,
N. elegans and N. miniatus share the same sequence while
N. serotinus falls into another branch.
The monotypic sect. Aurelia (N. broussonetii) has a chro-
mosomal structure close to N. tazetta, differing by two pairs
of chromosomes, which is consistent with its placement in the
ITS tree where it is nested within this group (Fig. 2A). In con-
trast, in the organellar DNA tree, N. broussonetii grouped with
N. elegans and N. serotinus (Fig. 1A). Consistent with this
finding, it presents isobrachial long chromosomes which are
also present in N. serotinus and some species of sect. Apodanthi
(Fernandes, 1940, 1975). There is considerable evidence for a
hybrid origin of N. broussonetii and our phylogenetic results
point towards N. tazetta s.l. and N. serotinus as possible parents.
Narcissus calcicola 1 (Electr. Suppl.: Table S3) represents
another likely case of hybridization or introgression, falling
within its own sect. Apodanthi in the organellar tree (Fig.
1A, group III) but together with N. gaditanus, within sect.
Jonquillae, in the ITS tree (Fig. 2A, group VI).
Divergence age estimates inferred from molecular dating.
— The divergence time estimates inferred from the chloro-
plast tree in BEAST suggest Neogene ages for the main inter-
nal nodes (Electr. Suppl.: Fig. S5). The early diversification
of Narcissus (13.80 Ma) occurred in the Middle Miocene in
the transition between the Langhian and Serravallian periods.
During this period at the end of the Middle Miocene Climatic
Optimum (MMCO) ca. 17–15 Ma, when the African and West
European plates were close, a drier and cooler climate was rap-
idly established that eliminated the dominance of subtropical
forests in the Mediterranean region. This event dramatically
affected the composition of western Mediterranean flora and
vegetation (Barrón & al., 2010) resulting in extinction of a large
number of Angiosperm lineages (Postigo Mijarra & al., 2009).
As a consequence of the climatic and biotic changes, the veg-
etation in the western Mediterranean (Iberian Peninsula, NW
Africa) evolved towards open grassland ecosystems (Wolfe,
1975; Zachos & al., 2001; Jiménez-Moreno & al., 2010). In
a climatic context of aridification, with dry summers and
cold winters coupled with bimodal rain seasons (spring and
autumn), associated with increasing numbers of grazing ungu-
late species in open woodlands (Domingo & al., 2014), the
adaptive advantage of geophytes, such as daffodils and other
monocots, is evident (Fragman & Shmida, 1997; Noy-Meir
& Oron, 2001). The same timing for the colonization of the
Mediterranean region, often from African sources, has been
inferred for other monocot groups characterized by geophytic
life-forms, like Androcymbium Willd. (Caujapé-Castells & al.,
2001) and subfamily Hyacinthoideae (Ali & al., 2013).
The distinct ages for the diversification of the two main
lineages of Narcissus (subg. Hermione, subg. Narcissus) suggest
different biogeographic scenarios. Our results date the early
diversification of subg. Narcissus in the mid-Tortonian (9.99 Ma)
with a subsequent diversification in the Iberian Peninsula
throughout the progressive onset of the Mediterranean climate.
Version of Record
Geologically, this took place after the break of the arc connect-
ing the East Iberian Peninsula and North Africa, before the final
Pliocene accretion of the Betic-Rif Cordilleras and the opening
of the Alboran Sea (Lonergan & White, 1997, Rosenbaum & al.,
2002). It is tempting to hypothesize that the predominant spring-
flowering in this subgenus evolved as an innovation within
the climatic context of increasing seasonality and progressive
onset of the Mediterranean climate. The autumnal flowering of
close relatives of Narcissus, such as those represented by three
of the four outgroups (Lapiedra martinezii, Nerine bowdenii,
Sternbergia lutea) and even species of Leucojum other than
L. aestivum as well as in other bulbs related to South African
lineages, makes this hypothesis conceivable.
In contrast, the age estimate for the crown node of subg.
Hermione is 5.98 Ma, which corresponds to diversification start-
ing in the late Messinian. During this period, successive sea
desiccation resulted in land connections between Mediterranean
islands, archipelagos and the mainland. This period has fre-
quently been invoked to explain lineage diversification, spe-
ciation and dispersal across the Mediterranean region, e.g.,
in Androcymbium (Caujapé-Castells & Jansen, 2003), Smilax
L. (Chen & al., 2014), Festuca subg. Schenodorus (P.Beauv.)
Peterm. (Inda & al., 2014). This scenario is consistent with the
widespread distribution of species of subg. Hermione in the
lowlands of the entire Mediterranean region, including some
islands, e.g., Malta, Crete, Cyprus (Turland & al., 1993; Mifsud
& Caruana, 2010; Hand & al., 2011) that were connected to the
mainland only at that time (Krijgsman & al., 1999).
Our mean age estimates are substantially more recent than
those obtained by Santos-Gally & al. (2012), i.e., 23.6, 21.6, 19.6
and 13.2 Ma for the stem ages of Narcissus and the crown ages
of Narcissus, as well as of subg. Narcissus and subg. Hermione,
respectively. Only in the first case (stem age of Narcissus),
their mean estimate falls within the 95% confidence intervals
provided here. The reasons for these discrepancies are likely the
calibration, ours being based on the most recent and complete
dating of angiosperm lineages published thus far (Magallón
& al., 2015), and on the speciation model prior, ours being a
birth-death process, which is more appropriate when using phy-
logenetically distant calibration points to account for lineage
extinction (Nee, 2006).
TAXONOMIC IMPLICATIONS
According to the results discussed above, we consider
that some taxonomic adjustments are needed to provide for-
mal recognition for those lineages that are strongly supported
in this study and by independent evidence, but that were not
recognized by most previous infrageneric classifications (see
in Electr. Suppl.: Fig. S9 how our schematic proposal matches
the phylogenetic results). In three cases (sect. Bulbocodii, sect.
Jonquillae, sect. Pseudonarcissi) these adjustments consist of
subdividing sections into two to reflect natural groups. Due to
the disparate placement of the resulting segregated groups in the
phylogenetic hypotheses generated, they cannot be placed to-
gether in a single section without introducing radical taxonomic
845
Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae)
changes in the infrageneric classification of the entire genus. In
comparison, the accommodation of lineages of hybrid origin in
the infrageneric treatment is more difficult, no matter whether
they are monophyletic in the two trees, since they could be
placed in the section corresponding to either the maternal or
paternal lineage. This is aggravated when the hybrid lineage
involves progenitors from the two subgenera, as in the allopoly-
ploids N. dubius and N. tortifolius. For this reason, we have been
conservative in maintaining those hybrid lineages as separate
sections, to reflect their dual placement and to preserve the
monophyly of the maternal and paternal sections. This is the
case for sect. Aurelia, sect. Dubii and sect. Ganymedes.
Another taxonomic change is needed because of the
placement of the type of the genus, N. poeticus and its relative
N. radiiflorus, in sect. Pseudonarcissi. This finding requires
changing the circumscription of sect. Narcissus to coincide
with what was, to date, sect. Pseudonarcissi. However, to main-
tain the recognition of the morphologically characteristic trum-
pet daffodils, we propose the recognition of two subsections
even if subsect. Pseudonarcissi is paraphyletic.
For subg. Hermione, we are proposing a conservative infra-
generic treatment because the eastern Mediterranean and further
eastern areas have not been sufficiently sampled for N. tazettae
and because of the suspected effects of allopolyploid taxa on
the phylogenetic trees. Further studies might eventually recom-
mend recognizing just a single section within subg. Hermione.
However, we prefer to be conservative at this point trying to rec-
oncile the evolution of the genus with morphological characters.
The placement of each species in the newly proposed infra-
generic taxonomic treatment can be seen below and the cor-
respondence between different treatments is shown in Electr.
Suppl.: Table S1 (see also Electr. Suppl.: Fig. S9).
Narcissus L., Sp. Pl.: 289. 1753 – Type N. poeticus L.
Narcissus subg. Narcissus
Nine sections are recognized:
1. Narcissus sect. Apodanthi A.Fern. in Bol. Soc. Brot., ser. 2,
40: 241. 1966 – Type N. rupicola Dufour.
Species included in our phylogeny. – Narcissus albimargi­
natus D.Müll.-Doblies & U.Müll.-Doblies, N. atlanticus Stern,
N. calcicola, N. cuatrecasasii Fern.Casas, N. cuatrecasasii
var. segimonensis Fern.Casas, N. marvieri Jahand. & Maire,
N. rupicola, N. scaberulus Henriq., N. watieri Maire.
2. Narcissus sect. Braxireon (Raf.) Valdés in Lagascalia 12:
274. 1984 – Type N. cavanillesii Barra & G.López.
Species included in our phylogeny. – Narcissus cavanillesii
(Cav.) Barra & G.López.
3. Narcissus sect. Bulbocodii DC. in Redouté, Liliac. 8: sub
t. 486. 1815. – Type N. bulbocodium L.
Species included in our phylogeny. – Narcissus bulboco-
dium, N. bulbocodium var. citrinus Baker, N. bulbocodium
subsp. graellsii, N. bulbocodium subsp. quintanillae*, N. bulbo­
codium var. nivalis (Graells) Baker, N. jacquemoudii Fern.Casas,
N. jeanmonodii Fern.Casas, N. tingitanus Fern.Casas.
846 Version of Record
TAXON 66 (4) • August 2017: 832–854
*  This taxon requires further work to determine its cor-
rect placement, although we include it here at this point since
most of our populations grouped with the remaining members
of this section.
4. Narcissus sect. Ganymedes (Salisb. ex Haw.) Schult. &
Schult.f. in Roemer & Schultes, Syst. Veg. 7: 952. 1830 –
Type N. triandrus L.
Species included in our phylogeny. – Narcissus triandrus,
N. triandrus var. loiseleurii (Rouy) A.Fern., N. triandrus subsp.
lusitanicus (Dorda Alcaraz & Fern.Casas) Barra, N. triandrus
subsp. pallidulus (Graells) Rivas Goday.
5. Narcissus sect. Jonquilla DC. in Redouté, Liliac. 8: sub
t. 486. 1815 – Type N. jonquilla L.
Species included in our phylogeny. – Narcissus cerrolazae,
N. cordubensis, N. fernandesii*, N. fernandesii var. major
A.Fern.*, N. jonquilla, N. jonquilla var. henriquesii Samp.,
N. viridiflorus, N. willkommii.
*  Further work is needed to reveal the hypothetical hybrid
origin of N. fernandesii (Blanchard, 1990) and the origin of the
tetraploid N. fernandesii var. major.
6. Narcissus sect. Juncifolii (A.Fern.) Zonn. in Pl. Syst. Evol.
275: 122. 2008 – Type N. assoanus Dufour.
Species included in our phylogeny. – Narcissus assoanus,
N. assoanus subsp. praelongus Barra & G.López, N. assoanus
subsp. rivas-martinezii (Fern.Casas) Barra, Díez & Ureña*,
N. baeticus*, N. gaditanus.
We recognize the division of sect. Jonquilllae into two
different sections based on cytological, genomic and our
own phylogenetic results. Zonneveld (2008) had already pro-
posed the placement of N. assoanus and N. gaditanus in sect.
Juncifolii due to substantial differences in genome size (< 20 pg
in sect. Juncifolii and > 30 pg in the remaining species of sect.
Jonquilllae). Zonneveld (2008) accepted the previous criterion
of Fernandes (1966, 1969) who split this section into two sub-
sections based on substantial karyotype differences.
*  Narcissus baeticus requires further work since, as dis-
cussed above, it grouped in sect. Juncifolii in the organellar tree
but fell into sect. Jonquillae in the ITS tree. Our phylogenetic
results support the recent transfer of subsp. rivas-martinezii
from N. fernandesii to N. assoanus (Barra & al., 2016) since
this subspecies fell within sect. Juncifolii in both the organellar
and the ITS trees.
7. Narcissus sect. Meridionalis I.Marques, Fuertes, Martins-
Loução, Moharrek & Nieto Fel., sect. nov. – Type: N. can-
tabricus DC.
Species included in our phylogeny. – Narcissus bulbo­co­
dium subsp. obesus*, N. cantabricus, N. hedraeanthus, N. per­
occidentalis Fern.Casas, N. romieuxii Braun-Blanq.
This newly described section encompasses several species
related to, and traditionally included in, sect. Bulbocodii, which
have here been revealed to constitute an independent lineage.
Morphologically, these species usually have creamy-yellow
to white concolor flowers, with usually exerted anthers and
TAXON 66 (4) • August 2017: 832–854 Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae)
some degree of zygomorphism, and are narrowly distributed
in mountainous regions in southern Iberia and Morocco. The
name Meridionalis (southern) refers to such a geographic pat-
tern compared to sect. Bulbocodii.
*  This taxon requires further study since different sam-
ples fell into the two hoop petticoat daffodils groups (sect.
Bulbocodii, sect. Meridionalis).
8. Narcissus sect. Narcissus
8.1 Narcissus subsect. Narcissus
Species included in our phylogeny. – Narcissus poeticus,
N. radiiflorus.
The inclusion of this group within sect. Pseudonarcissi
had already been detected in the phylogenetic study of Graham
& Barrett (2004). Fernandes (1968) had also remarked on the
similarity of karyotypes. The genome size value measured in
N. poeticus (26 pg) is also within the range of values reported
for sect. Pseudonarcissi by Zonneveld (2008) who also stated
that “it is the deviating morphology of the flower, with a very
short and orange coloured corona, that has prevented so far to
unite these sections”. Based on all evidence, we now include it
as subsection of Narcissus.
8.2 Narcissus subsect. Pseudonarcissi (DC.) I.Marques,
Fuertes, Martins-Loução, Moharrek & Nieto Fel., stat.
nov. ≡ Narcissus sect. Pseudonarcissi DC. in Redouté,
Liliac. 8: sub t. 486. 1815 – Type N. pseudonarcissus DC.
Species included in our phylogeny. – Narcissus abscissus
(Haw.) Schult. & Schult.f., N. alpestris, N. asturiensis, N. bicolor,
N. bujei*, N. confusus Pugsley, N. cyclamineus, N. hispanicus,
N. hispanicus subsp. perez-chiscanoi (Fern.Casas) Fern.Casas,
N. jacetanus Fern.Casas, N. macrolobus (Jord.) Pugsley, N. mi-
nor L., N. moleroi Fern.Casas, N. nanus (Haw.) Spach, N. nobilis,
N. nobilis var. leonensis, N. pallidiflorus Pugsley, N. primigenius
(Fern. Suárez ex Laínz) Fern.Casas & Laínz, N. pseudonarcis-
sus, N. pseudonarcissus subsp. eugeniae (Fern.Casas) Fern.
Casas, N. pseudonarcissus subsp. munozii-garmendiae (Fern.
Casas) Fern.Casas, N. portensis Pugsley, N. pumilus Salisb.
*  Further work is necessary to determine its correct
placement.
9. Narcissus sect. Nevadensis Zonn. in Pl. Syst. Evol. 275: 125.
2008 – Type N. nevadensis Pugsley.
Species included in our phylogeny. – Narcissus alcaracen-
sis, N. longispathus, N. nevadensis, N. segurensis, N. yepesii
Our results support the earlier subdivision of sect.
Pseudonarcissi proposed by Zonneveld (2008) into his new
sect. Nevadensis based on substantial genome size differ-
ences (30–39 pg in sect. Nevadensis vs. 22–26 pg in sect.
Pseudonarcissi with the exception of N. bicolor that has 68 pg).
Narcissus subg. Hermione (Haw.) Spach
Five sections are recognized:
1. Narcissus sect. Angustifolii (A.Fern.) Fern.Casas in Anales
Jard. Bot. Madrid 55: 174. 1997 – Type N. elegans (Haw.)
Spach.
Version of Record
Species included in our phylogeny. – Narcissus elegans,
N. miniatus.
This group was created by Abilio Fernandes as a mono-
typic subsection to accommodate the tetraploid N. elegans,
which he thought to be an ancient tetraploid that should be
differentiated from N. serotinus and the remaining species
of section Tazettae. Our phylogenetic results reveal different
placements for N. elegans and N. serotinus in the two trees
(Fig. 1A, 2A) especially in the analyses without allopolyploids
(Electr. Suppl.: Fig. S7, S8). Therefore, we think it is appropriate
to follow a conservative criterion and recognize Fernandes’s
section including the allopolyploid derivative N. miniatus.
2. Narcissus sect. Aurelia (Gay) Baker, Handb Amaryll.: 2.
1888 – Type N. broussonetii Lagasca.
Species included in our phylogeny. – Narcissus broussonetii.
In addition to the cytological and phylogenetic evidence for
a hybrid origin, this monotypic section has a distinct morpho-
logical feature: a vestigial corona with exerted anthers whose
filaments are attached to the floral tube just below the tepals
(Blanchard, 1990).
3. Narcissus sect. Dubii Fern.Casas in Fontqueria 6: 39, 41.
1984 – Type: N. dubius Gouan.
Species included in our phylogeny. – Narcissus dubius,
N. tortifolius.
4. Narcissus sect. Serotini Parl., Fl. Ital. 3: 157. 1858 – Type
N. serotinus L.
Species included in our phylogeny. – Narcissus serotinus.
5. Narcissus sect. Tazetta DC. in Redouté, Liliac. 8: sub t. 486.
1815 – Type N. tazetta L.
Species included in our phylogeny. – Narcissus pachybol-
bus Durieu, N. pannizianus Parl., N. papyraceus, N. polyanthus
Loisel., N. tazetta*.
*  As stated above, a deeper study of this species is needed
to fully understand its variation.
ACKNOWLEDGEMENTS
The authors wish to dedicate this paper to the memory of Abílio
Fernandes† and Rosette Batarda†, whose pioneer work established
the grounds for evolutionary studies in Narcissus. The authors
thank A. González Fernández de Castro, A. Susanna, A. Tribsch, Á.
Bueno, C. Herrera, C. Sérgio, C. Garcia, C. Lefebvre, C. Tauleigne-
Gomes, D. Draper, E. Triano†, M. Ruiz, F. Domínguez, G. Kamari, G.
Bacchetta, H. Nava, J.C. Moreno, J. Feise, J. Mota, J. Pedrol, J. Pérez,
J. Paiva, M. Luceño, M. Vincens, M. Lizana, M. Serrano, M. Medrano,
N. Membrives, P. Escobar, V. Valcárcel, and S. Jury for supplying
material, the curator staff of herbarium LISU, and I. Sanmartín for
guidance on the dating analyses. The authors also thank the helpful
comments of four reviewers, M. Pirie and J. Kadereit, who have con-
tributed substantially to improve the manuscript, and to Franz Stadler
for his excellent editing of this complex manuscript. This work has
been supported by a FPVI European-funded Integrated Infrastructure
847
Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae)
Initiative “SYNTHESYS“ to IM (ES-TAF 023-2004), FCT, Ministério
de Ciência e Tecnologia, Portugal (PD/1245/2004), plus grants from the
CSIC-Comunidad de Madrid, Spain, (CCG07-CSIC/AMB-1978), from
the Spanish Ministerio de Ciencia y Tecnología (CGL2007-66516).
LITERATURE CITED
Aedo, C. 2013. Narcissus L. Pp. 340–397 in: Rico, E., Crespo, M.B.,
Quintanar, A., Herrero. A. & Aedo, C. (eds.), Flora iberica.
Madrid: Real Jardín Botánico-CSIC.
Ali, S.S., Pfosser, M., Wetschnig, W., Martínez-Azorín, M., Crespo,
M.B. & Yu, Y. 2013. Out of Africa: Miocene dispersal, vicariance,
and extinction within Hyacinthaceae subfamily Urgineoideae.
J. Integr. Pl. Biol. 55: 950–964. https://doi.org/10.1111/jipb.12065
Alvarez, I. & Wendel, J.F. 2003. Ribosomal ITS sequences and plant
phylogenetic inference. Molec. Phylogen. Evol. 29: 417–434.
https://doi.org/10.1016/S1055-7903(03)00208-2
Arroyo, J. & Dafni, A. 1995. Variations in habitat, season, flower traits
and pollinators in dimorphic Narcissus tazetta L. (Amaryllidaceae)
in Israel. New Phytol. 129: 135–146.
https://doi.org/10.1111/j.1469-8137.1995.tb03017.x
Bailey, C.D., Carr, T.G., Harris, S.A. & Hughes, C.E. 2003.
Characterization of angiosperm nrDNA polymorphism, paralogy,
and pseudogenes. Molec. Phylogen. Evol. 29: 435–455.
https://doi.org/10.1016/j.ympev.2003.08.021
Baker, H.G. 1964. Variation in style length in relation to outbreeding
in Mirabilis (Nyctaginaceae). Evolution 18: 507–512.
https://doi.org/10.1111/j.1558-5646.1964.tb01627.x
Baker, J.G. 1875. Review of the genus Narcissus. Pp. 61–86 in:
Burbidge, F.W., The Narcissus: Its history and culture; with
Coloured plates and descriptions of all known species and prin-
cipal varieties. London: L. Reeve.
https://books.google.vg/books?id=UxdBAAAAYAAJ
Barra, A. 1999. Notas sobre Narcissus L. (Amaryllidaceae). Anales
Jard. Bot. Madrid 57: 178–180.
Barra, A. & López González, G. 1982. Una nueva localidad para
Narcissus tortifolius Fdez. Casas. Anales Jard. Bot. Madrid 39: 212.
Barra, A., Diez Dominguez, R. & Ureña Plaza, F. 2016. Novedades
en Narcissus L. (Amaryllidaceae). Flora Montiber. 63: 103–111.
Barrett, S.C.H. & Harder, L.D. 2005. The evolution of polymorphic
sexual systems in daffodils (Narcissus). New Phytol. 165: 45–53.
https://doi.org/10.1111/j.1469-8137.2004.01183.x
Barrett, S.C.H., Lloyd, D.G. & Arroyo, J. 1996. Stylar polymorphisms
and the evolution of heterostyly in Narcissus (Amaryllidaceae). Pp.
339–376 in: Lloyd, D.G. & Barrett, S.C.H. (eds.), Floral biology:
Studies on floral evolution in animal-pollinated plants. New York:
Chapman & Hall. https://doi.org/10.1007/978-1-4613-1165-2_13
Barrett, S.C.H., Cole, W.W. & Herrera, C.M. 2004. Mating patterns
and genetic diversity in the wild daffodil Narcissus longispathus
(Amaryllidaceae). Heredity 92: 459–465.
https://doi.org/10.1038/sj.hdy.6800441
Barrón, E., Rivas-Carballo, R., Postigo-Mijarra, J.M., Alcalde-
Olivares, C., Vieira, M., Castro, L., Pais, J. & Valle-Hernández,
M. 2010. The Cenozoic vegetation of the Iberian Peninsula: A syn-
thesis. Rev. Palaeobot. Palynol. 162: 382–402.
https://doi.org/10.1016/j.revpalbo.2009.11.007
Bateman, A.J. 1952. Trimorphism and self-incompatibility in
Narcissus. Nature 170: 496–497. https://doi.org/10.1038/170496d0
Berkov, S., Martínez-Francés, V., Bastida, J., Codina, C. & Ríos,
S. 2014. Evolution of alkaloid biosynthesis in the genus Narcissus.
Phytochemistry 99: 95–106.
https://doi.org/10.1016/j.phytochem.2013.11.002
Blanchard, J.W. 1990. Narcissus: A guide to wild daffodils. Surrey:
Alpine Garden Society.
Caujapé-Castells, J. & Jansen, R.K. 2003. The influence of the
848 Version of Record
TAXON 66 (4) • August 2017: 832–854
Miocene Mediterranean desiccation on the geographical expan-
sion and genetic variation of Androcymbium gramineum (Cav.)
McBride (Colchicaceae). Molec. Ecol. 12: 1515–1525.
https://doi.org/10.1046/j.1365-294X.2003.01839.x
Caujapé-Castells, J., Jansen, R.K., Membrives, N., Pedrola-
Monfort, J., Montserrat, J.M. & Ardanuy, A. 2001. Historical
biogeography of Androcymbium Willd. (Colchicaceae) in Africa:
Evidence from cpDNA RFLPs. Bot. J. Linn. Soc. 136: 379–392.
https://doi.org/10.1111/j.1095-8339.2001.tb00581.x
Chen, C., Qi, Z.C., Xu, X.H., Comes, H.P., Koch, M.A., Jin, X.J.,
Fu, C.X. & Qiu, Y.X. 2014. Understanding the formation of
Mediterranean–African–Asian disjunctions: Evidence for Miocene
climate-driven vicariance and recent long-distance dispersal in
the Tertiary relict Smilax aspera (Smilacaceae). New Phytol. 204:
243–255. https://doi.org/10.1111/nph.12910
Clement, M., Posada, D. & Crandall, K.A. 2000. TCS: A computer
program to estimate gene genealogies. Molec. Ecol. 9: 1657–1659.
https://doi.org/10.1046/j.1365-294x.2000.01020.x
Díaz Lifante, Z., Andrés, C., Viruel, J. & Cabrera, A. 2009. The allo-
polyploid origin of Narcissus obsoletus (Alliaceae): Identification
of parental genomes by karyotype characterization and genomic
in situ hybridation. Bot. J. Linn. Soc. 159: 477–498.
https://doi.org/10.1111/j.1095-8339.2009.00951.x
Domingo, M.S., Badgley, C., Azanza, B., DeMiguel, D. & Alberdi,
M.T. 2014. Diversification of mammals from the Miocene of Spain.
Paleobiology 40: 197–221. https://doi.org/10.1666/13043
Drummond, A.J., Suchard, M.A., Xie, D. & Rambaut, A. 2012.
Bayesian phylogenetics with BEAUti and the BEAST 1.7. Molec.
Biol. Evol. 29: 1969–1973. https://doi.org/10.1093/molbev/mss075
Escudero, M., Eaton, D.A., Hahn, M. & Hipp, A.L. 2014. Genotyping-
by-sequencing as a tool to infer phylogeny and ancestral hybridiza-
tion: A case study in Carex (Cyperaceae). Molec. Phylogen. Evol.
79: 359–367. https://doi.org/10.1016/j.ympev.2014.06.026
Fernandes, A. 1934. Nouvelles études caryologiques sur le genre
Narcissus L. Bol. Soc. Brot., ser. 2, 9: 1–198.
Fernandes, A. 1935. Remarque sur l’hétérostylie de Narcissus triandrus
et de N. reflexus. Bol. Soc. Brot., ser. 2, 10: 5–15.
Fernandes, A. 1937. Sur l’origine du Narcissus dubius Gouan. Bol. Soc.
Brot., ser. 2, 12: 93–118.
Fernandes, A.1939. Sur l’origine du Narcissus jonquilloides Willk.
Sci. Genet. 1: 16–61.
Fernandes, A. 1940. Sur la position systématique et l’origine de
Narcissus broussonettii Lag. Bol. Soc. Brot., ser. 2, 14: 53–66.
Fernándes, A. 1951. Sur la phylogénie des espèces du genre Narcissus.
Bol. Soc. Brot., ser. 2, 25: 113–192.
Fernandes, A. 1964. Contribution à la connaissance de la génétique de
l´hétérostylie chez le genre Narcissus L.I. Résultats de quelques
croisements. Bol. Soc. Brot., ser. 2, 38: 81–96.
Fernandes, A.1966. Nouvelles études caryologiques sur la section Jon­
quilla DC. du genre Narcissus L. Bol. Soc. Brot., ser. 2, 40: 207–261.
Fernandes, A. 1967. Contribution à la connaisance de la biosystema-
tique de quelques espèces du genre Narcissus L. Portugaliae Acta
Biol., Sér. B, Sist. 9: 1–44.
Fernandes, A. 1968. Keys to the identification of native and naturalized
taxa of the genus Narcissus L. Daffodil Tulip Year Book 1968: 37–66.
Fernandes, A. 1969. Contribution to the knowledge of the biosyste-
matics of some species of genus Narcissus L. Pp. 245–284 in:
V Simposio de Flora Europaea (20–30 de mayo de 1967). Sevilla:
Universidad de Sevilla.
http://bibdigital.rjb.csic.es/ing/Libro.php?Libro=7185
Fernandes, A. 1975. L’evolution chez le genre Narcissus L. Anales Inst.
Bot. Cavanilles 32: 843–872.
Fernandes, A. & Neves, J.B. 1941. Sur l’origine des formes de
Narcissus bulbocodium L. à 26 chromosomes. Bol. Soc. Brot.,
ser. 2, 15: 43–132.
Fernández Casas, F.J. 1983. Materiales para uma monografia de
Narcissus L. Fontqueria 3: 23–32.
TAXON 66 (4) • August 2017: 832–854 Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae)
Fernández Casas, F.J. 1984. Remiendas y enmiendas en el género
Narcissus L. Fontqueria 6: 35–50.
Fragman, O. & Shmida, A. 1997. Diversity and adaptation of wild geo-
phytes along an aridity gradient in Israel. Acta Hort. 430: 795–802.
https://doi.org/10.17660/ActaHortic.1997.430.127
Graham, S.W. & Barrett, S.C.H. 2004. Phylogenetic reconstruc-
tion of the evolution of stylar polymorphisms in Narcissus
(Amaryllidaceae). Amer. J. Bot. 91: 1007–1021.
https://doi.org/10.3732/ajb.91.7.1007
Hall, T.A. 1999. BioEdit: A user-friendly biological sequence alignment
editor and analysis. http://www.mbio.ncsu.edu/BioEdit/bioedit.html
Hand, R., Hadjikyriakou, G.N. & Christodoulou, C.S. 2011– (con-
tinuously updated). Flora of Cyprus – A dynamic checklist. http://
www.flora-of-cyprus.eu/ (accessed 17 Sep 2015).
Hanks, G.R. 2002. Commercial production of Narcissus bulbs. Pp
53–130 in: Hanks, G.R. (ed.), Narcissus and daffodil: The genus
Narcissus. New York: Taylor and Francis.
Hao, W. & Golding, G.B. 2006. Asymmetrical evolution of cyto­chrome
bd subunits. J. Molec. Evol. 62: 132–142.
https://doi.org/10.1007/s00239-005-0005-7
Haworth, A.H. 1831. Narcissinearum monographia, ed. 2. London:
Ridgway. http://bibdigital.rjb.csic.es/ing/Libro.php?Libro=7195
Heinrich, M. & Teoh, H.L. 2004. Galanthamine from snowdrop—The
development of a modern drug against Alzheimer’s disease from
local Caucasian knowledge. J. Ethnopharmacol. 92: 147–162.
https://doi.org/10.1016/j.jep.2004.02.012
Heled, J., Bryant, D. & Drummond, A.J. 2013. Simulating gene trees
under the multispecies coalescent and time-dependent migration.
B. M. C. Evol. Biol. 13: 44. https://doi.org/10.1186/1471-2148-13-44
Herbert, W. 1837. Amaryllidaceae: preceded by an Attempt to arrange
the monocotyledonous orders; and followed by a Treatise on cross-
bred vegetables, and supplement. London: Ridgway.
https://doi.org/10.5962/bhl.title.116651
Huelsenbeck, J.P. & Ronquist, F. 2001. MRBAYES. Bayesian infer-
ence of phylogeny. Bioinformatics 17: 754–755.
https://doi.org/10.1093/bioinformatics/17.8.754
Inda, L.A., Sanmartín, I., Buerki, S. & Catalán, P. 2014. Mediter­
ranean origin and Miocene–Holocene Old World diversification of
meadow fescues and ryegrasses (Festuca subgenus Sche­do­norus and
Lolium). J. Biogeogr. 41: 600–614. https://doi.org/10.1111/jbi.12211
Ito, M., Kawamoto, A., Kita, Y., Yukawa, T. & Kurita, S. 1999.
Phylogenetic relationships of Amaryllidaceae based on matK
sequence data. J. Pl. Res. 112: 207–216.
https://doi.org/10.1007/PL00013874
Jiménez, J.F., Sánchez-Gómez, P., Guerra, J., Molins, A. & Rosselló,
J.A. 2009. Regional speciation or taxonomic inflation? The sta-
tus of several narrowly distributed and endangered species of
Narcissus using ISSR and nuclear ribosomal ITS markers. Folia
Geobot. 44: 145–158. https://doi.org/10.1007/s12224-009-9040-2
Jiménez-Moreno, G., Suc, J.P. & Fauquette, S. 2010. Mio-Pliocene
vegetation and climate estimates from the Iberian Peninsula de-
duced from pollen data. Rev. Palaeobot. Palynol. 162: 403–415.
https://doi.org/10.1016/j.revpalbo.2009.08.001
Joly, S., McLenachan, P.A. & Lockhart, P.J. 2009. A statistical ap-
proach for distinguishing hybridization and incomplete lineage sort-
ing. Amer. Naturalist 174: e54–e70. https://doi.org/10.1086/600082
Jones, K.E., Reyes-Betancort, J.A., Hiscock, S.J. & Carine, M.A.
2014. Allopatric diversification, multiple habitat shifts, and hybrid-
ization in the evolution of Pericallis (Asteraceae), a Macaronesian
endemic genus. Amer. J. Bot. 101: 637–651.
https://doi.org/10.3732/ajb.1300390
Kington, S. 2008. The International Daffodil Register and Classified
List. United Kingdom: Royal Horticultural Society.
Koopowitz, H., Hawkins, B.A., Van der Hoven, H., Donnison-
Morgan, D. & Howe, M. 2002. Comparisons between the repro-
ductive fitness of two species of Narcissus L. (Amaryllidaceae)
from Southern Spain. Herbertia 56(2001): 67–76.
Version of Record
Krijgsman, W., Hilgen, F.J., Raffi, I., Sierro, F.J. & Wilson, D.S.
1999. Chronology, causes and progression of the Messinian salinity
crisis. Nature 400: 652–655. https://doi.org/10.1038/23231
Kubatko, L.S., Carstens, B.C. & Knowles, L.L. 2009. STEM: Species
Tree Estimation using Maximum likelihood for gene trees under
coalescence. Bioinformatics 25: 971–973.
https://doi.org/10.1093/bioinformatics/btp079
Librado, P. & Rozas, J. 2009. DnaSP v5: A software for comprehensive
analysis of DNA polymorphism data. Bioinformatics 25: 1451–
1452. https://doi.org/10.1093/bioinformatics/btp187
Lledó, M.D., Davis, A.P, Crespo, M.B, Chase, M.W. & Fay,
M.F. 2004. Phylogenetic analysis of Leucojum and Galanthus
(Amaryllidaceae) based on plastid matK and nuclear ribosomal
spacer (ITS) DNA sequences and morphology. Pl. Syst. Evol. 246:
223–243. https://doi.org/10.1007/s00606-004-0152-0
Lonergan, L. & White, N. 1997. Origin of the Betic-Rif mountain belt.
Tectonics 16: 504–522. https://doi.org/10.1029/96TC03937
Magallón, S., Gómez-Acevedo, S., Sánchez-Reyes, L.L. & Hernández-​
Hernández, T. 2015. A metacalibrated time-tree documents the
early rise of flowering plant phylogenetic diversity. New Phytol.
207: 437–453. https://doi.org/10.1111/nph.13264
Marques, I., Rosselló-Graell, A., Draper, D. & Iriondo, J.M. 2007.
Pollination patterns limit hybridization between two sympatric
species of Narcissus (Amaryllidaceae). Amer. J. Bot. 94: 1352–1359.
https://doi.org/10.3732/ajb.94.8.1352
Marques, I., Nieto Feliner, G., Martins-Loução, M.A. & Fuertes
Aguilar, J. 2008. Cytoplasm and nuclear ribosomal DNA evi-
dence reveal a differentiated southern Iberian lineage of wild trum-
pets (Narcissus sect. Pseudonarcissus) [Poster Abstract]. P. 80
in: Xth Symposium of the International Organization of Plant
Biosystematists, 2–4 July, 2008, Vysoké Tatry, Slovakia.
http://www.iopb.org/PDF/BOOK%20OF%20ABSTRACTS.pdf
Marques, I., Nieto Feliner, G., Draper, D., Martins-Loução, M.A.
& Fuertes Aguilar, J. 2010. Unraveling cryptic reticulate rela-
tionships and the origin of orphan hybrid disjunct populations in
Narcissus. Evolution 64: 2353–2368.
Marques, I., Nieto Feliner, G., Martins-Loução, M.A. & Fuertes
Aguilar, J. 2011. Fitness in Narcissus hybrids: Low fertility is
overcome by early hybrid vigor, absence of exogenous selection
and high bulb propagation. J. Ecol. 99: 1508–1519.
https://doi.org/10.1111/j.1365-2745.2011.01876.x
Marques, I., Nieto Feliner, G., Martins-Loução, M.A. & Fuertes
Aguilar, J. 2012. Genome size and base composition variation
in natural and experimental Narcissus (Amaryllidaceae) hybrids.
Ann. Bot. (Oxford) 109: 257–264. http://doi.org/10.1093/aob/mcr282
Marques, I., Fuertes Aguilar, J., Martins-Loução, M.A. & Nieto
Feliner, G. 2012b. Spatial-temporal patterns of flowering asyn-
chrony and pollinator fidelity in hybridizing species of Narcissus.
Evol. Ecol. 26: 1433–1450.
https://doi.org/10.1007/s10682-012-9554-6
Marques, I., Jurgens, A., Fuertes Aguilar, J. & Nieto Feliner, G.
2016. Convergent recruitment of new pollinators is triggered by
independent hybridization events in Narcissus. New Phytol. 210:
731–742. https://doi.org/10.1111/nph.13805
Mathew, B. 2002. Classification of the genus Narcissus. Pp. 30–52 in:
Hanks, G.R. (ed.), Narcissus and daffodil: The genus Narcissus.
London: Taylor and Francis.
Maureira-Butler, I.J., Pfeil, B.E., Muangprom, A., Osborn, T.C. &
Doyle, J.J. 2008. The reticulate history of Medicago (Fabaceae).
Syst. Biol. 57: 466–482. https://doi.org/10.1080/10635150802172168
Mayol, M. & Rosselló, J.A. 2001. Why nuclear ribosomal DNA spacers
(ITS) tell different stories in Quercus. Molec. Phylogen. Evol. 19:
167–176. https://doi.org/10.1006/mpev.2001.0934
Medrano, M., Herrera, C.M. & Barrett, S.C.H. 2005. Herkogamy and
mating patterns in the self-compatible daffodil Narcissus longi­
spathus. Ann. Bot. (Oxford) 95: 1105–1111.
https://doi.org/10.1093/aob/mci129
849
Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae)
Medrano, M., López-Perea, E. & Herrera, C.M. 2014. Population
genetics methods applied to a species delimitation problem:
Endemic trumpet daffodils (Narcissus section Pseudonarcissi)
from the southern Iberian Peninsula, Int. J. Pl. Sci. 175: 501–517.
https://doi.org/10.1086/675977
Meerow, A.W. & Snijman, D.A. 2006. The never ending story:
Multigene approaches to the phylogeny of Amaryllidaceae, and
assessing its familial limits. Aliso 22: 353–364.
Meerow, A.W., Fay, M.F, Guy, C.L, Li, Q.-B., Zaman, F.Q. & Chase,
M.W. 1999. Systematics of Amaryllidaceae based on cladistic
analysis of plastid rbcL and trnL-F sequence data. Amer. J. Bot.
86: 1325–1345. https://doi.org/10.2307/2656780
Meerow, A.W., Francisco-Ortega, J., Kuhn, D.N. & Schnell, R.J.
2006. Phylogenetic relationships and biogeography within the
Eurasian clade of Amaryllidaceae based on plastid ndhF and
nrDNA ITS sequences: Lineage sorting in a reticulate area? Syst.
Bot. 31: 42–60. https://doi.org/10.1600/036364406775971787
Mifsud, S. & Caruana, E. 2010. Records of Narcissus elegans (fam.
Amaryllidaceae) and notes on the wild Narcissus in the Maltese
islands. Centr. Medit. Naturalist 5: 19–29.
Mort, M.E., Soltis, P.S, Soltis, D.E. & Mabry, M. 2000. Comparison of
three methods for estimating internal support on phylogenetic trees.
Syst. Biol. 49: 160–171. https://doi.org/10.1080/10635150050207456
Nee, S. 2006. Birth-death models in macroevolution. Ann. Rev. Ecol.
Evol. Syst. 37: 1–17.
https://doi.org/10.1146/annurev.ecolsys.37.091305.110035
Nieto Feliner, G. & Rosselló, J.A. 2012. Concerted evolution of multi­
gene families and homoeologous recombination. Pp. 171–193 in:
Wendel, J.F., Greilhuber, J., Dolezel, J. & Leitch, I.J. (eds.), Plant
genome diversity, vol. 1. Vienna: Springer.
https://doi.org/10.1007/978-3-7091-1130-7_12
Noy-Meir, I. & Oron, T. 2001. Effects of grazing on geophytes in
Mediterranean vegetation. J. Veg. Sci. 12: 749–760.
https://doi.org/10.2307/3236862
Nylander, J.A. 2002. MrModeltest, version 1.1b. http://www.ebc.uu.se/
systzoo/staff/nylander.html.
Payseur, B.A. & Rieseberg, L.H. 2016. A genomic perspective on
hybridization and speciation. Molec. Ecol. 25: 2337–2360.
https://doi.org/10.1111/mec.13557
Pelser, P.B., Kennedy, A.H., Tepe, E.J., Shidler, J.B., Nordenstam,
B., Kadereit, J.W. & Watson, L.E. 2010. Patterns and causes of in-
congruence between plastid and nuclear Senecioneae (Asteraceae)
phylogenies. Amer. J. Bot 97: 856–873.
https://doi.org/10.3732/ajb.0900287
Pérez, R., Vargas, P. & Arroyo, J. 2004. Convergent evolution of
flower polymorphism in Narcissus (Amaryllidaceae). New Phytol.
161: 235–252. https://doi.org/10.1046/j.1469-8137.2003.00955.x
Pérez-Barrales, R., Vargas, P. & Arroyo, J. 2006. New evidence for
the Darwinian hypothesis of heterostyly: Breeding systems and
pollinators in Narcissus sect. Apodanthi. New Phytol. 171: 553–567.
https://doi.org/10.1111/j.1469-8137.2006.01819.x
Phytos, D. & Kamari, G. 1974. Zytotaxonomische Beitrage zur Flora
von Kreta. Bot. Not. 127: 302–308.
Posada, D. & Crandall, K.A. 2002. The effect of recombination on the
accuracy of phylogeny reconstruction. J. Molec. Evol. 54: 396–402.
https://doi.org/10.1007/s00239-001-0034-9
Postigo Mijarra, J.M., Barrón, E., Gómez Manzaneque, F. & Morla,
C. 2009. Floristic changes in the Iberian Peninsula and Balearic
Islands (south-west Europe) during the Cenozoic. J Biogeogr. 36:
2025–2043. https://doi.org/10.1111/j.1365-2699.2009.02142.x
Rambaut, A. & Drummond, A.J. 2007. Tracer, version 1.4. http://
beast.bio.ed.ac.uk/Tracer
Rambaut, A, Suchard, M.A., Xie, D. & Drummond, A.J. 2014.
Tracer, version 1.6. http://beast.bio.ed.ac.uk/Tracer
RHS (Royal Horticultural Society) 2016. Botanical classifica-
tion, updated Dec 2016. https://www.rhs.org.uk/plants/pdfs/
plant-registration-forms/daffbotanical
850 Version of Record
TAXON 66 (4) • August 2017: 832–854
Romero, A.T., Sánchez Castillo, P.M. & Ruiz Rejón, M. 1983.
Sobre cariología, morfología y corología de Narcissus tortifolius
Fernández Casas. Fontqueria 4: 7–10.
Ronquist, F., Teslenko, M., Van der Mark, P., Ayres, D.L.,
Darling, A., Höhna, S., Larget, B., Liu, L., Suchard, M.A. &
Huelsenbeck, J.P. 2012. MrBayes 3.2: Efficient Bayesian phyloge-
netic inference and model choice across a large model space. Syst.
Biol. 61: 539–542. https://doi.org/10.1093/sysbio/sys029
Rønsted, N., Savolainen, V., Mølgaard, P. & Jäger, A.K. 2008.
Phylogenetic selection of Narcissus species for drug discovery.
Biochem. Syst. Ecol. 36: 417–422.
https://doi.org/10.1016/j.bse.2007.12.010
Rosenbaum, G., Lister, G.S. & Duboz, C. 2002. Reconstruction of
the tectonic evolution of the western Mediterranean since the
Oligocene. J. Virtual Explor. 8: 107–130.
https://doi.org/10.3809/jvirtex.2002.00053
Sánchez Gómez, P., Carrión Vilches, M.A., Carrillo, A.F.,
Hernández, A., Jiménez, J.F. & Rabal, G. 2000. Dos plantas
de interés para la flora de Murcia. Anales Jard Bot. Madrid 58:
198–199.
Santos-Gally, R., Vargas, P. & Arroyo, J. 2012. Insights into Neogene
Mediterranean biogeography based on phylogenetic relationships
of mountain and lowland lineages of Narcissus (Amaryllidaceae).
J. Biogeogr. 39: 782–789.
https://doi.org/10.1111/j.1365-2699.2011.02526.x
Schierup, M.H. & Hein, J. 2000a. Consequences of recombination on
traditional phylogenetic analysis. Genetics 156: 879–891.
Schierup, M.H. & Hein, J. 2000b. Recombination and the molecular
clock. Molec. Biol. Evol. 17: 1578–1579.
https://doi.org/10.1093/oxfordjournals.molbev.a026256
Spach, E. 1846. Histoire naturelle des végétaux: Phanérogames, vol.
12. Paris: Librairie Encyclopedique de Roret.
https://doi.org/10.5962/bhl.title.44839
Swofford, D.L. 1999. PAUP* 4.0: Phylogenetic analysis using parsi-
mony (and *other methods). Sunderland, Massachusetts: Sinauer.
Syvanen, M. 2002. Rates of ribosomal RNA evolution are uniquely
accelerated in eukaryotes. J. Molec. Evol. 55: 85–91.
https://doi.org/10.1007/s00239-001-2306-9
Templeton, A.R., Crandall, K.A. & Sing, C.F. 1992. A cladistic
analysis of phenotypic associations with haplotypes inferred from
restriction endonuclease mapping and DNA sequence data. III.
Cladogram estimation. Genetics 132: 619–633.
Turland, N.J., Chilton, L. & Press, J.R. 1993. Flora of the Cretan Area:
Annotated checklist and atlas. London: Natural History Museum.
Vrondeli, A., Kefalas, P. & Kokkalou, E. 2005. A new alkaloid from
Narcissus serotinus L. Pharmazie 60: 559–560.
https://doi.org/10.1002/chin.200546187
Webb, D.A. 1980. Narcissus. Pp. 78–84 in: Tutin, T.G., Heywood, V.H.,
Burges, N.A., Moore, D.M., Valentine, D.H., Walters, S.M. &
Webb, D.A. (eds.), Flora Europaea, vol. 5. Cambridge: Cambridge
University Press.
Wendel, J.F. & Doyle, J.J. 1998. Phylogenetic incongruence: Window
into genome history and molecular evolution. Pp. 265–296 in:
Soltis, P.S. & Doyle, J. (eds.), Molecular systematics of plants II.
Boston: Kluwer. https://doi.org/10.1007/978-1-4615-5419-6_10
Wolfe, J.A. 1975. Some aspects of plant geography of the Northern
Hemisphere during the Late Cretaceous and Tertiary. Ann.
Missouri. Bot. Gard. 62: 264–279. https://doi.org/10.2307/2395198
Yang, Z. 1996. Among-site rate variation and its impact on phylogenetic
analysis. Trends Ecol. Evol. 11: 367–373.
https://doi.org/10.1016/0169-5347(96)10041-0
Zachos, J., Pagani, M., Sloan, L., Thomas, E. & Billups, K. 2001.
Trends, rhythms, and aberration in global climate 65 Ma to pres-
ent. Science 292: 686–693. https://doi.org/10.1126/science.1059412
Zachos, J.C., Dickens, G.R. & Zeeb, R.E. 2008. An early Cenozoic
perspective on green-house warming and carbon-cycle dynamics.
Nature 451: 279–283. https://doi.org/10.1038/nature06588
TAXON 66 (4) • August 2017: 832–854 Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae)

Zonneveld, B. 2008. The systematic value of nuclear DNA content for Zonneveld, B. 2010. The involvement of Narcissus hispanicus Gouan
all species of Narcissus L. (Amaryllidaceae). Pl. Syst. Evol. 275: in the origin of Narcissus bujei and of cultivated trumpet daffodils
109–132. https://doi.org/10.1007/s00606-008-0015-1 (Amaryllidaceae). Anales Jard. Bot. Madrid 67: 29–39.

Appendix 1. Taxa sampled arranged according to the traditional infrageneric classification largely based on Fernandes (1968), including populations sampled,
chromosome numbers (2n) counted in this study (B, presence of B chromosome; *, new count for the taxon), number of individuals used to count chromosomes
(# indiv. cytogenetics), number of individuals sampled for chloroplast and mitochondrial regions (# indiv. orgDNA), number of non-redundant orgDNA sequences
used in the phylogenetic analysis (# non-red. seq. orgDNA), number of individuals sampled for ITS (# indiv. ITS clones), number of clones retrieved per individual
(# ITS clones), number of non-redundant clones used in the phylogenetic analyses (# non-red. ITS clones), and GenBank accession numbers (all accession
numbers not ending with “.1” represent new sequences). Codes: CBNB, Conservatoire Botanique National de Brest; COI, Jardim Botânico de Coimbra, Abilio
Fernandes Collection; BCN: Jardí Botànic de Barcelona; MIM, Jardí Botànic Marimurtra; JB-MNHN, Jardim Botanico Lisboa-Museu Nacional de Historia
Natural; JBS, Jardí Botànic de Sóller; RJB, Real Jardín Botánico Madrid, CSIC; RNG, The University of Reading, Herbarium RNG.
Taxon, population, voucher/provenience of sample, GenBank accession numbers (ndhF, matK, cob, atpA, ITS)
Section Apodanthi: N. albimarginatus D Müll.-Doblies & U. Müll.-Doblies, N. albimarginatus 1, Morocco, Chaouen, Jbel Bouhachem, pers. coll. A. Mohammed,
KT210953, KR872971, KT272903, KT279959, KY971398; N. albimarginatus 2, Morocco, Chaouen, Jbel Kelti, pers. coll. A. Mohammed, KT210954, KR872972,
KT272904, KT279960, KY971399; N. atlanticus Stern, N. atlanticus, Morocco, Oukaimeden, pers. coll. A. Mohammed, KT863444, KT033140, KT808873, KT878479,
KY971412; N. calcicola Mendonça, N. calcicola 1, Portugal, Serra de Aire e Candeeiros, Marques 1908, Flores & Draper (LISU 188516), KT340508, KT021558,
KT316751, KT594671, KY971400; N. calcicola 2, Portugal, Rocha da Pena, Marques 1907, Brehm & Draper (LISU 188473), KT340509, KT021559, KT316752,
KT594672, KY971401; N. calcicola 3, Portugal, Serra de Monte Figo, C. Garcia s.n., leaf sample provided by C. Garcia, KT340510, KT021560, KT316753, KT594673,
KY971402; N. ×carringtonii Rozeira, N. ×carringtonii (scaberulus × triandrus subsp. pallidulus), Portugal, Ervedal da Beira, Marques 799 & Draper, pers. coll.
A. Fernandes, KT799083, KT033095, KT808828, KT878434, KY971407; N. cuatrecasasii Fern.Casas & al., N. cuatrecasasii 1, Spain, Mágina, pers. coll. E. Triano,
KT340521, KT021571, KT316764, KT594684, KY971408; N. cuatrecasasii 2, Spain, Sierra de Cazorla, Medrano & Herrera 2006-01, leaf sample provided by
C. Herrera, KT340520, KT021570, KT316763, KT594683, KY971409; N. cuatrecasasii var. segimonensis Fern.Casas, N. cuatrecasasii var. segimonensis, Spain,
Quesada, Puerto de Tiscar, pers. coll. E. Triano, KT340519, KT021569, KT316762, KT594682, KY971410; N. cuatrecasasii Fern.Casas & al. × N. triandrus L.,
N. cuatrecasasii × N. triandrus, Spain, Mágina, pers. coll. E. Triano, KT340522, KT021572, KT316765, KT594685, KY971411; N. marvieri Jahand. & Maire,
N. marvieri 1, Morocco, Jbel Tazzeka, pers. coll. A. Mohammed, KT799111, KT021617, KT619090, KT594730, KY971414; N. marvieri 2, Morocco, Jbel Zerhoun,
pers. coll. A. Mohammed, KT799112, KT021618, KT619091, KT594731, –; N. rupicola Dufour, N. rupicola 1, Spain, Puerto de Canencia, Marques 1982 & Draper
(LISU 265718), KT799076, KT033088, KT808821, KT878427, KY971415; N. rupicola 2, Portugal, Manteigas, Penhas Douradas, Marques 1880 & Draper (LISU
265716), KT799077, KT033089, KT808822, KT878428, KY971416; N. rupicola 3, Portugal, Manteigas, Penhas Saude, Marques 1881 & Draper (LISU 265717),
KT799078, KT033090, KT808823, KT878429, –; N. rupicola 4, Spain, Puerto de Guadarrama, Marques 1978 & Draper (LISU 265719), KT799079, KT033091,
KT808824, KT878430, –; N. rupicola 5, Spain, Sierra Segura, leaf sample provided by D. Draper, KT799081, KT033093, KT808826, KT878432, –; N. rupicola
Dufour × N. triandrus L., N. rupicola × N. triandrus, Spain, Puerto de Guadarrama, Marques 808 & Draper (LISU 265743), KT799080, KT033092, KT808825,
KT878431, KY971417; N. scaberulus Henriq, N. scaberulus 1, Portugal, Tranvancinha, Ervedal da Beira, Marques 1985 & Draper (LISU 265720), KT799082,
KT033094, KT808827, KT878433, KY971405; N. scaberulus 2, Portugal, Oliveira do Conde, Marques 1984 & Draper (LISU 265721), KT799084, KT033096,
KT808829, KT878435, KY971406; N. watieri Maire, N. watieri 1, Morocco, Marrakesh, Oukaimeden, pers. coll. A. Mohammed, KT863443, KT033139, KT808872,
KT878478, KY971413; N. watieri 2, Morocco, Marrakesh, Tizi’n’Test, pers. coll. A. Mohammed, KT863442, KT033138, KT808871, KT878477, –; — Section
Aurelia: N. broussonetii Lag., N. broussonetii 1, Morocco, El Jadida, living coll. of JB-MNHN 2080, KT124406/​KT124407, KR872993/​K R872994, KT272925/​
KT272926, KT340488/​KT340489, KY930351; N. broussonetii 2, Morocco, Essaouira, living coll. of JB-MNHN 2081, KT124408/​KT124409, KR872997/​K R872998,
KT272929/​KT272930, KT340492/​KT340493, KY930352; N. broussonetii 3, Morocco, Safi, pers. coll. A. Mohammed, KT124410/​KT124411, KR872995/​K R872996,
KT272927/​KT272928, KT340490/​KT340491, –; — Section Braxireon: N. ×alentejanus Fern.Casas, N. ×alentejanus (cavanillesii × serotinus), Portugal, Ajuda,
Marques 1158 & Draper (LISU 181995), HM371834.1, KM349526.1, HM231770.1, HM174663.1, KY971378; N. cavanillesii Barra & G.López, N. cavanillesii 1,
Portugal, Ajuda, Marques 1155 & Draper (LISU 188451), HM371429.1, KR872965, HM231365.1, HM174258.1, KY971373; N. cavanillesii 2, Spain, Facinas, Marques
1166 & Draper (LISU 187635), HM371534.1, KM349505.1, HM231470.1, HM174363.1, KY971374; N. cavanillesii 3, Morocco, Oued el Malah, Tetouan, living coll.
of JB-MNHN 1191, HM371729.1, KM349519.1, HM231665.1, HM174558.1, KY971376; N. cavanillesii 4, Spain, Posadas, Marques 1144 & Draper (LISU 187628),
HM371619.1, KM349509.1, HM231555.1, HM174448.1, KY971375; N. cavanillesii 5, Morocco, Souk Khémis des Anjra, living coll. of JB-MNHN 1204, HM371789.1,
KM349523.1, HM231725.1, HM174618.1, KY971377; N. cavanillesii 6, Morocco, Sidi-el-Yamani, living coll. of JB-MNHN 1187, HM371804.1, KM349524.1,
HM231740.1, HM174633.1, –; N. ×perezlarae Font Quer, N. ×perezlarae (cavanillesii × miniatus), Spain, Oliva, Marques 1174 & Draper, living coll. of JB-MNHN
1174, HM371924.1, KM349538.1, HM231860.1, HM174753.1, MF039090; — Section Bulbocodii: N. ×barrae Fern.Casas, N. ×barrae (bulbocodium × cantabricus),
Spain, Sierra de Garza, living coll. of RJB, KT327006, KT210992, KT304245, KT594649, KY992372; N. ×brevitubulosus A.Fern., N. ×brevitubulosus (asturiensis
× bulbocodium), Portugal, Lagoa Comprida, Marques 2083 & Draper, pers. coll. A. Fernandes, KT799130, KT021636, KT619109, KT594749, KY992384; N. bul­
bocodium L., N. bulbocodium 1, Morocco, Agadir, F.L. Domínguez s.n., leaf sample provided by F.L. Domínguez, KT326987, KT210973, KT304226, KT594630,
KY992438; N. bulbocodium 2, Portugal, Santarem, Marques 1873 & Draper (LISU 265723), KT326986, KT210972, KT304225, KT594629, KY992380; N. bulboco-
dium 3, Portugal, Montes Juntos, Marques 1884 & Draper (LISU 187646), KT326988, KT210974, KT304227, KT594631, –; N. bulbocodium 4, Portugal, Beja,
Marques 1878 & Draper (LISU 187434), KT326989, KT210975, KT304228, KT594632, –; N. bulbocodium 5, Portugal, Aldeia da Luz, Marques 1876 & Draper
(LISU 187436), KT326990, KT210976, KT304229, KT594633, –; N. bulbocodium 6, Portugal, Mourao, Marques 1875 & Draper (LISU 187672), KT326991, KT210977,
KT304230, KT594634, –; N. bulbocodium 7, Portugal. COI s.n., pers. coll. A. Fernandes, KT326992, KT210978, KT304231, KT594635, –; N. bulbocodium 8, Spain,
Gijón, H. Nava & A. Bueno s.n., leaf sample provided by H. Nava, KT326993, KT210979, KT304232, KT594636, –; N. bulbocodium 9, Spain, Gijón, H. Nava &
A. Bueno s.n., leaf sample provided by H. Nava, KT326994, KT210980, KT304233, KT594637, –; N. bulbocodium 10, Spain, Sierra de Garza, Marques 2072, leaf
sample provided by J.C. Moreno, KT327007, KT210993, KT304246, KT594650, –; N. bulbocodium 11, Portugal, Mogador, Marques 1890, leaf sample provided by
A. Rosselló-Graell, KT327008, KT210994, KT304247, KT594651, –; N. bulbocodium 12, Portugal, Outeiro do Pombo, Marques 1885 (LISU 265724), KT326995,
KT210981, KT304234, KT594638, –; N. bulbocodium 13, Portugal, Penhas Douradas, Marques 2041 & Draper (LISU 265725), KT326985, KT210971, KT304224,
KT594628, –; N. bulbocodium 14, Spain, Arroyo Cuncos, Marques 2029 (LISU 188402), KT326996, KT210982, KT304235, KT594639, –; N. bulbocodium 15,
Portugal, Ponte da Ajuda, Marques 1874 & Draper (LISU 188505), KT326997, KT210983, KT304236, KT594640, –; N. bulbocodium 16, Spain, Puerto del Tremedal,
living coll. of RJB, KT326998, KT210984, KT304237, KT594641, –; N. bulbocodium 17, Portugal, S. Marcos, Marques 203 (LISU 188486), KT326999, KT210985,
KT304238, KT594642, –; N. bulbocodium 18, Portugal, Ribeira de Godelim, Marques 1881 & Draper (LISU 188491), KT327000, KT210986, KT304239, KT594643,
–; N. bulbocodium 19, Spain, Ames, Ponte Maceira, M. Serrano s.n., leaf sample provided by M. Serrano, KT327001, KT210987, KT304240, KT594644, –; N. bul-
bocodium 20, Spain, Santíz, M. Lizana s.n., leaf sample provided by M. Lizana, KT327002, KT210988, KT304241, KT594645, –; N. bulbocodium 21, Spain, Pico
Cervero, M. Lizana s.n., leaf sample provided by M. Lizana, KT327003, KT210989, KT304242, KT594646, –; N. bulbocodium 22, Portugal, Magoito, Marques 2040
& Draper (LISU 188402), KT327010, KT210996, KT304249, KT594653, –; N. bulbocodium 23, Portugal, living coll. of JB-MNHN s.n., KT327004/​KT327005,
KT210990/​KT210991, KT304243/​KT304244, KT594647/​KT594648, –; N. bulbocodium 24, Spain, Sierra de Urbasa, J.C. Moreno s.n., leaf sample provided by J.C.
Moreno, KT327009, KT210995, KT304248, KT594652, –; N. bulbocodium var. citrinus Baker, N. bulbocodium var. citrinus, Spain, Puerto de Las Señales, A. Bueno

Version of Record 851

Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae) TAXON 66 (4) • August 2017: 832–854

Appendix 1. Continued.
s.n., leaf sample provided by A. Bueno, KT327020, KT211006, KT304259, KT594663, KY992371; N. bulbocodium subsp. graellsii (Webb ex Graells) K.Richt.,
N. bulbocodium subsp. graellsii, Spain, Puerto de Guadarrama, Marques 1947 & Draper (LISU 265726), KT326981, KT210967, KT279997, KT340506, KY992377;
N. bulbocodium var. nivalis (Graells) Baker, N. bulbocodium var. nivalis 1, Spain, Puerto de Las Señales, H. Nava & A. Bueno s.n., leaf sample provided by H. Nava,
KT799128, KT021634, KT619107, KT594747, KY992381; N. bulbocodium var. nivalis 2, Spain, Puerto de Tarna, H. Nava & A. Bueno s.n., leaf sample provided by
H. Nava, KT799129, KT021635, KT619108, KT594748, –; N. bulbocodium subsp. obesus (Salisb.) Maire, N. bulbocodium subsp. obesus 1, Portugal, Alte, Marques
1901 & Draper (LISU 188088), KT327013, KT210999, KT304252, KT594656, KY992365/​KY992366; N. bulbocodium subsp. obesus 2, Portugal, Penhas Douradas,
Marques 1903 & Draper (LISU 265727), KT327011, KT210997, KT304250, KT594654, KY992367; N. bulbocodium subsp. obesus 3, Portugal, Portela, C. Sergio
21611, pers. coll. C. Sérgio, KT327012, KT210998, KT304251, KT594655, –; N. bulbocodium subsp. obesus 4, Portugal, Berlenga, Cristina Gomes s.n., pers. coll.
C. Gomes, KT327024, KT211010, KT304263, KT594667, –; N. bulbocodium subsp. obesus 5, Portugal, Cape Espichel, Marques 1897 & Draper (LISU 265728),
KT327014, KT211000, KT304253, KT594657, –; N. bulbocodium subsp. obesus 6, Portugal, Lagoa Comprida, Marques 1898 & Draper (LISU 265729), KT327015,
KT211001, KT304254, KT594658, –; N. bulbocodium subsp. obesus 7, Portugal, Magoito, Cristina Gomes s.n., pers. coll. C. Gomes, KT327016, KT211002, KT304255,
KT594659, –; N. bulbocodium subsp. obesus 8, Morocco, Oukaimeden, S.L. Jury 8945. RNG s.n., leaf sample provided by S.L. Jury, KT327018, KT211004, KT304257,
KT594661, –; N. bulbocodium subsp. obesus 9, Portugal, Arrábida, Marques 1894 & Draper (LISU 265730), KT327021, KT211007, KT304260, KT594664, –;
N. bulbocodium subsp. obesus 10, Portugal. S Arrábida, Marques 1896 & Draper (LISU 265731), KT327022, KT211008, KT304261, KT594665, –; N. bulbocodium
subsp. obesus 11, Portugal, Rocha da Pena, Marques 1900 & Draper (LISU 188475), KT327025, KT211011, KT304264, KT594668, –; N. bulbocodium subsp. obe-
sus 12, Portugal, Zimbral, Marques 1902 & Draper (LISU 188088), KT327023, KT211009, KT304262, KT594666, –; N. bulbocodium subsp. obesus (Salisb.) Maire
× N. calcicola Mendonça, N. bulbocodium subsp. obesus × N. calcicola, Portugal, Rocha da Pena, Marques 1970 & Draper (LISU 265745), KT327026, KT211012,
KT304265, KT594669, KY992368; N. bulbocodium subsp. quintanilhae A.Fern., N. bulbocodium subsp. quintanilhae 1, Portugal, Cerdeira, Marques 1903 & Draper
(LISU 265732), KT326983, KT210969, KT304222, KT594626, KY992369; N. bulbocodium subsp. quintanilhae 2, Portugal, Figueira de Castelo Rodrigo, Marques
1904 & Draper (LISU 265733), KT326984, KT210970, KT304223, KT594627, KY992370; N. cantabricus DC., N. cantabricus 1, Spain, Puerto Virgen. MIM98110734,
living coll. of MIM, KT340511, KT021561, KT316754, KT594674, KY992374; N. cantabricus 2, Spain, Sierra de Garza, leaf sample provided by J. Nieto, KT340513,
KT021563, KT316756, KT594676, KY992376; N. cantabricus 3, Spain, Mérida, germplasm coll. of BCN, KT340512, KT021562, KT316755, KT594675, –; N. can-
tabricus 4, Spain, Granada, Sierra Nevada, El Serrallo, J.B. Hoya de Pedraza s.n., J.B. Hoya de Pedraza, KT340514, KT021564, KT316757, KT594677, –; N. ×cazor­
lanus Fern.Casas, N. ×cazorlanus (hedraeanthus × triandrus subsp. pallidulus), living coll. of JB-MNHN s.n., KT340540, KT021594, KT316787, KT594707,
KY992373; N. ×felineri Fern.Casas, N. ×felineri (bulbocodium × primigenius), Spain, Puerto de Tarna, H. Nava & A. Bueno s.n., leaf sample provided by H. Nava,
KT799131, KT021637, KT619110, KT594750, KY992383; N. hedraeanthus (Webb & Heldr.) Colmeiro, N. hedraeanthus 1, Spain, Sierra de Segura, Collado de
Gontar, Medrano & Herrera 2006-07, leaf sample provided by C. Herrera, KT340538, KT021592, KT316785, KT594705, KY992360; N. hedraeanthus 2, Spain,
Sierra de Cazorla, La Cabrilla, Medrano & Herrera 2006-08, leaf sample provided by C. Herrera, KT340539, KT021593, KT316786, KT594706, –; N. jacquemoudii
Fern.Casas, N. jacquemoudii, Morocco, Tahnaout, pers. coll. A. Mohammed, KT326982, KT210968, KT304221, KT594625, KY992378; N. jeanmonodii Fern.Casas,
N. jeanmonodii, Morocco, Oukaimeden, pers. coll. A. Mohammed, KT327017, KT211003, KT304256, KT594660, KY992382; N. peroccidentalis Fern.Casas,
N. peroccidentalis, Morocco, Kénitra, pers. coll. A. Mohammed, KT799132, KT033069, KT808802, KT594764, KY992363; N. romieuxii Braun-Blanq. & Maire,
N. romieuxii 1, Morocco, Chaouen, pers. coll. A. Mohammed, KT799072, KT033084, KT808817, KT878423, KY992364; N. romieuxii 2, Morocco, Ketama, pers.
coll. A. Mohammed, KT799073, KT033085, KT808818, KT878424, –; N. tenuifolius Salisb., N. tenuifolius, BCN920158, germplasm coll. of BCN, KT340507,
KT211013, KT304266, KT594670, KY992362; N. tingitanus Fern.Casas, N. tingitanus, Morocco, Tanger, living coll. of JB-MNHN 2076, KT327019, KT211005,
KT304258, KT594662, KY992379; N. ×tuckeri Barra & G.López, N. ×tuckeri 1 (  fernandesii × hedraeanthus), BCN930526, germplasm coll. of BCN, KT340541,
KT021595, KT316788, KT594708, KY992361; N. ×tuckeri 2 (  fernandesii × hedraeanthus), living coll. of JB-MNHN s.n., KT340542, KT021596, KT316789,
KT594709,  –; — Section Ganymedes: N. ×rozeirae Pérez-Chisc. & Fernández Casas, N. ×rozeirae (bulbocodium × triandrus subsp. pallidulus), BCN920211,
germplasm coll. of BCN, KT799075, KT033087, KT808820, KT878426, KY992389/​KY992437; N. ×susannae Fern.Casas, N. ×susannae (cantabricus × triandrus
subsp. pallidulus), Spain, Mérida, living coll. of RJB, KT799086, KT033098, KT808831, KT878437, KY992375; N. triandrus L., N. triandrus 1, Spain, La Cabra,
E. Triano s.n., pers. coll. E. Triano, KT799092, KT033104, KT808837, KT878443, KY992385; N. triandrus 2, Portugal, Ervedal da Beira, living coll. of JB-MNHN
2036, KT799096, KT033108, KT808841, KT878447, KY992386; N. triandrus 3, Spain, Gijon, H. Nava & A. Bueno s.n., leaf sample provided by H. Nava, KT799109,
KT033121, KT808854, KT878460, KY992387; N. triandrus 4, Spain, Gijon, H. Nava & A. Bueno s.n., leaf sample provided by H. Nava, KT799110, KT033122,
KT808855, KT878461, –; N. triandrus 5, Spain, Sierra de Mágina, Marques 2037, pers. coll. E. Triano, KT799093, KT033105, KT808838, KT878444, –; N. trian-
drus 6, Spain, Sierra de Segura, Navalcaballo, Medrano & Herrera 2006-05, leaf sample provided by C. Herrera, KT799094, KT033106, KT808839, KT878445, –;
N. triandrus 7, Spain, Sierra de Cazorla, Vadillo-Castril, Medrano & Herrera 2006-06, leaf sample provided by C. Herrera, KT799095, KT033107, KT808840,
KT878446, –; N. triandrus var. loiseleurii (Rouy) A.Fern., N. triandrus var. loiseleurii, France, Glénan. CBNB s.n., leaf sample provided by CBNB, KT799108,
KT033120, KT808853, KT878459, KY992388; N. triandrus subsp. lusitanicus (Dorda Alcaraz & Fern.Casas) Barra, N. triandrus subsp. lusitanicus 1, Portugal,
Águas Belas, Marques 1969 & Draper (LISU 182000), KT799097, KT033109, KT808842, KT878448, KY992390; N. triandrus subsp. lusitanicus 2, Spain. BCN920167,
germplasm coll. of BCN, KT799098, KT033110, KT808843, KT878449, KY992391; N. triandrus subsp. lusitanicus 3, Portugal, Dornes, Marques 1968 & Draper
(LISU 182001), KT799099, KT033111, KT808844, KT878450, KY992392; N. triandrus subsp. lusitanicus 4, Portugal, Mogadouro, Marques 2038 (LISU 265734),
KT799100, KT033112, KT808845, KT878451, –; N. triandrus subsp. pallidulus (Graells) Rivas Goday, N. triandrus subsp. pallidulus 1, Spain, Sierra de Mágina,
Marques 2039, pers. coll. E. Triano, KT799106, KT033118, KT808851, KT878457, KY992394; N. triandrus subsp. pallidulus 2, Portugal, Miranda do Douro, Marques
2014 (LISU 265735), KT799107, KT033119, KT808852, KT878458, KY992395; N. triandrus subsp. pallidulus 3, Portugal, Penamacor, C. Garcia s.n., leaf sample
provided by C. Garcia, KT799103, KT033115, KT808848, KT878454, –; N. triandrus subsp. pallidulus 4, Portugal, Penhas da Saúde, Marques 2010 & Draper (LISU
265736), KT799104, KT033116, KT808849, KT878455, –; N. triandrus subsp. pallidulus 5, Spain, Puerto de Guadarrama, Marques 2008 & Draper (LISU 265737),
KT799102, KT033114, KT808847, KT878453, –; N. triandrus subsp. pallidulus 6, Portugal, Penhas da Saúde, Torre, Marques 2009 & Draper (LISU 265738),
KT799105, KT033117, KT808850, KT878456, –; N. triandrus L. × N. bulbocodium L., N. triandrus × N. bulbocodium, Portugal, Mogadouro, Marques 2015 &
Draper (LISU 265746), KT799101, KT033113, KT808846, KT878452, KY992393; — Section Jonquillae: N. ×alleniae Donnison-Morgan, N. ×alleniae 1 (serotinus
× viridiflorus), Spain, Medina-Sidonia, living coll. of JB-MNHN 1173, KT326957/​KT326958, KR872966/​K R872967, KT272898, KT279954/​KT279955, KY971427/​
KY971391; N. ×alleniae 2 (serotinus × viridiflorus), Spain, San José Valle. (LISU 181997), KT326959, KR872968, KT272900, KT279956, –; N. assoanus Dufour
ex Schult. & Schult.f., N. assoanus 1, Spain, Coll Baix, germplasm coll. of BCN, KT210960, KR872978, KT272910, KT279966, KY971423; N. assoanus 2, Spain,
Granadella, Marques 1848, Soler & Draper, pers. coll. J. Soler, KT210961, KR872979, KT272911, KT279967, KY971424; N. assoanus 3, Spain, Montgó, Marques
1847, Soler & Draper, pers. coll. J. Soler, KT210962, KR872980, KT272912, KT279968, –; N. assoanus 4, Spain, Montserrat, Marques 2221 & Draper, germplasm
coll. of BCN, KT210963, KR872981, KT272913, KT279969, –; N. assoanus subsp. praelongus Barra & G.López, N. assoanus subsp. praelongus, Spain, Grazalema,
Puerto de las Palomas, E. Triano s.n., pers. coll. E. Triano, KT210964, KR872982, KT272914, KT279970, KY971425; N. assoanus subsp. rivas-martinezii (Fern.
Casas) Barra, Díez & Ureña, N. assoanus subsp. rivas-martinezii, Spain, Málaga, Sierra de las Nieves, E. Triano s.n., pers. coll. E. Triano, KT210965, KR872983,
KT272915, KT279971, KY971426; N. baeticus Fern.Casas, N. baeticus, Spain, Carcabuey, E. Triano s.n., pers. coll. E. Triano, KT326969, KR872999, KT279985,
KT340494, KY971438; N. cerrolazae Ureña, N. cerrolazae, Spain, Montecorto, Ronda, E. Triano s.n., pers. coll. E. Triano, KT340515, KT021565, KT316758,
KT594678, KY971436; N. cordubensis Fern.Casas, N. cordubensis, Spain, Benamejí, E. Triano s.n., pers. coll. E. Triano, KT340518, KT021568, KT316761, KT594681,
KY971437; N. fernandesii Pedro, N. fernandesii 1, Portugal, Vale da Pedra, Marques 1941 & Draper (LISU 188015), KT340529, KT021583, KT316776, KT594696,
KY971440; N. fernandesii 2, Spain, Escañuela. MIM98110716, living coll. of MIM, KT340532, KT021586, KT316779, KT594699, KY971441; N. fernandesii 3,
Spain, Lucena, E. Triano s.n., pers. coll. E. Triano, KT340533, KT021587, KT316780, KT594700, –; N. fernandesii 4, Spain, Santuario Virgen de la Cabeza, E. Triano

852 Version of Record

TAXON 66 (4) • August 2017: 832–854 Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae)

Appendix 1. Continued.
s.n., pers. coll. E. Triano, KT340531, KT021585, KT316778, KT594698, –; N. fernandesii var. major A.Fern., N. fernandesii var. major, Portugal, Vale da Pedra,
Marques 1939 & Draper (LISU 188016), KT340516, KT021566, KT316759, KT594679, KY971442; N. fernandesii Pedro × N. papyraceus Ker Gawl., N. fernandesii
× N. papyraceus, Portugal, Perto da Quinta do Vale da Pedra, Marques 1946 & Draper (LISU 265744), KT340530, KT021584, KT316777, KT594697, KY971443;
N. gaditanus Boiss. & Reut., N. gaditanus 1, Portugal, Benafim, Marques 1946 (LISU 188477), KT340534, KT021588, KT316781, KT594701, KY971403; N. gadi-
tanus 2, Spain, Ojén, Juánar, leaf sample provided by P. Escobar, KT340537, KT021591, KT316784, KT594704, KY971404; N. gaditanus 3, Portugal, Rocha da Pena,
Marques 1945 (LISU 188474), KT340535, KT021589, KT316782, KT594702, –; N. gaditanus 4, Portugal, Serra Monte Figo, C. Garcia s.n., leaf sample provided by
C. Garcia, KT340536, KT021590, KT316783, KT594703, –; N. jonquilla L., N. jonquilla 1, Portugal, Serra da Estrela. MIM98110716, living coll. of MIM, KT304268,
KT021602, KT316795, KT594715, KY971433; N. jonquilla 2, Spain, La Nava, leaf sample provided by P. Escobar, KT304269, KT021603, KT316796, KT594716,
KY971434; N. jonquilla 3, Portugal, Ponte da Ajuda, Marques 1953, Rosselló-Graell (LISU 188492), KT304270, KT021604, KT316797, KT594717, –; N. jonquilla 4,
Portugal, Pulo do Lobo, Marques 1956 (LISU 265722), KT304272, KT021606, KT316799, KT594719, –; N. jonquilla 5, Portugal, Mourao, Rosselló-Graell 1567
(LISU 187647), KT304273, KT021607, KT316800, KT594720, –; N. jonquilla 6, Portugal, Alandroal, Rosselló-Graell 1568 (LISU 187650), KT304274, KT021608,
KT316801, KT594721, –; N. jonquilla 7, Portugal, Aldeia da Luz, Rosselló-Graell 1569 (LISU 187647), KT304275, KT021609, KT316802, KT594722, –; N. jonquilla 8,
Spain, Arroyo de Cuncos, Marques 1957 & Draper (LISU 188507), KT304267, KT021601, KT316794, KT594714, –; N. jonquilla var. henriquesii Samp., N. jonquilla
var. henriquesii, Portugal, Ponte da Ajuda, Marques 1958, Rosselló-Graell (LISU 188500), KT304271, KT021605, KT316798, KT594718, KY971435; N. viridiflorus
Schousb., N. viridiflorus 1, Spain. BCN930017, germplasm coll. of BCN, KT863432, KT033128, KT808861, KT878467, KY971428; N. viridiflorus 2, Morocco, Cape
Spartel, living coll. of JB-MNHN 1208, KT863433, KT033129, KT808862, KT878468, KY971429; N. viridiflorus 3, Morocco, Kenitra, pers. coll. A. Mohammed,
KT863434, KT033130, KT808863, KT878469, KY971430; N. viridiflorus 4, Spain, La Línea, Marques 1138 & Draper, living coll. of JB-MNHN 1138, KT863436,
KT033132, KT808865, KT878471, KY971431; N. viridiflorus 5, Spain, Medina-Sidonia, Marques 1172 & Draper, living coll. of JB-MNHN 1172, KT863437,
KT033133, KT808866, KT878472, KY971432; N. viridiflorus 6, Morocco, Salé, pers. coll. A. Mohammed, KT863435, KT033131, KT808864, KT878470, –;
N. willkommii (Samp.) A.Fern., N. willkommii 1, Portugal, Alvor, Paiva 1528, pers. coll. J. Paiva, KT863438, KT033134, KT808867, KT878473, KY971439; N. willkom-
mii 2, Portugal. COI s.n., pers. coll. A. Fernandes, KT863440, KT033136, KT808869, KT878475, –; N. willkommii 3, Portugal, Loulé, Marques 2028 & Draper, pers.
coll. A. Fernandes, KT863439, KT033135, KT808868, KT878474, –; N. willkommii 4, Portugal, Ribeira da Quarteira, Marques 2027 & Draper (LISU 188478),
KT863441, KT033137, KT808870, KT878476, –; — Section Narcissus: N. poeticus L., N. poeticus 1, Andorra, Coll de Ordino, J. Pedrol 7941, pers. coll. J. Pedrol,
KT124416, KT033071, KT808804, KT594766, KY992421; N. poeticus 2, BCN890125, germplasm coll. of BCN, KT124417, KT033072, KT808805, KT594767, –;
N. radiiflorus Salisb., N. radiiflorus, Austria, Steiermark, Enns Valley, A. Tribsch s.n., leaf sample provided by A. Tribsch, KT124419, KT033083, KT808816,
KT878422, KY992420; — Section Pseudonarcissus: N. abscissus (Haw.) Schult. & Schult.f., N. abscissus 1, BCN920441, germplasm coll. of BCN, KT210951,
KR872969, KT272901, KT279957, KY992439; N. abscissus 2, Spain, Valle de Aran, germplasm coll. of BCN, KT210952, KR872970, KT272902, KT279958, –;
N. alcaracensis Ríos, D.Rivera, Alcaraz & Obón, N. alcaracensis 1, Spain, Peñascosa, Laguna Fuente Raja, Medrano & Herrera 2006-03, leaf sample provided by
C. Herrera, KT210955, KR872973, KT272905, KT279961, KY992401; N. alcaracensis 2, Spain, Peñascosa, Medrano & Herrera 2005-07, leaf sample provided by
C. Herrera, KT210956, KR872974, KT272906, KT279962, –; N. alcaracensis 3, Spain, Peñascosa, Rio Pesebre, Medrano & Herrera 2006-04, leaf sample provided
by C. Herrera, KT210957, KR872975, KT272907, KT279963, –; N. alpestris Pugsley, N. alpestris, Spain, San Juan de la Peña, Marques 2049, leaf sample provided
by L. Villar, KT210958, KR872976, KT272908, KT279964, KY992440; N. asturiensis (Jordan) Pugsley, N. asturiensis 1, Spain, Escorial, F.L. Dominguez s. n., leaf
sample provided by F.L. Domínguez, KT326960, KR872984, KT272916, KT279972, KY992423; N. asturiensis 2, Spain, Lago Isoba, H. Nava & A. Bueno s.n., leaf
sample provided by A. Bueno, KT326966, KR872990, KT272922, KT279978, KY992424; N. asturiensis 3, Portugal, Lagoa Comprida, Marques 1856 & Draper
(LISU 265739), KT326961, KR872985, KT272917, KT279973, –; N. asturiensis 4, Portugal, Penhas Douradas, Marques 1850 & Draper (LISU 265740), KT326963,
KR872987, KT272919, KT279975, –; N. asturiensis 5, Portugal, Manteigas, Marques 1852 & Draper (LISU 265741), KT326964, KR872988, KT272920, KT279976,
–; N. asturiensis 6, Spain, Somiedo, M. Serrano s.n., leaf sample provided by M. Serrano, KT326965, KR872989, KT272921, KT279977, –; N. asturiensis 7, Portugal,
S. Lagoa Comprida, Marques 1857 & Draper (LISU 265742), KT326967, KR872991, KT272923, KT279979, –; N. asturiensis 8, Spain, Sierra de Urbasa, J.C. Moreno
s.n., leaf sample provided by J.C. Moreno, KT326968, KR872992, KT272924, KT279980, –; N. asturiensis (Jordan) Pugsley × N.  bulbo­codium subsp. obesus
(Salisb.) Maire, N. asturiensis × N. bulbocodium subsp. obesus, Portugal, Lagoa Comprida. (LISU 265747), KT326962, KR872986, KT272918, KT279974, KY992422;
N. bicolor L., N. bicolor 1, Spain, Portalet, L. Villar s.n., leaf sample provided by L. Villar, KT326971, KR873001, KT279987, KT340496, KY992412; N. bicolor 2,
Spain, Valle de Aran, germplasm coll. of BCN, KT326970, KR873000, KT279986, KT340495, –; N. bujei Fern.Casas, N. bujei 1, Spain, Bienservida, Medrano &
Herrera 2006-13, leaf sample provided by C. Herrera, KT326972, KR873002, KT279988, KT340497, KY992404; N. bujei 2, Spain, Charco Negro, Medrano &
Herrera s.n., leaf sample provided by C. Herrera, KT326979, KR873009, KT279995, KT340504, KY992405; N. bujei 3, Spain, Collado Zamora, Medrano & Herrera
2005-09, leaf sample provided by C. Herrera, KT326973, KR873003, KT279989, KT340498, –; N. bujei 4, Spain, Sierra de Cazorla, El Chorro, Medrano & Herrera
2006-11, leaf sample provided by C. Herrera, KT326974, KR873004, KT279990, KT340499, –; N. bujei 5, Spain, Torredelcampo, El Megatín, Medrano & Herrera
2006-12, leaf sample provided by C. Herrera, KT326975, KR873005, KT279991, KT340500, –; N. bujei 6, Spain, Sierra de las Nieves, Los Quejigales, Medrano &
Herrera 2006-14, leaf sample provided by C. Herrera, KT326976, KR873006, KT279992, KT340501, –; N. bujei 7, Spain, Sierra de Mágina, Puerto de la Mata,
Medrano & Herrera 2005-08, leaf sample provided by C. Herrera, KT326977, KR873007, KT279993, KT340502, –; N. bujei 8, Spain, Sierra de las Nieves, E. Triano
s.n., pers. coll. E. Triano, KT326978, KR873008, KT279994, KT340503, –; N. confusus Pugsley, N. confusus, Spain, Cañamero., germplasm coll. of BCN, KT340517,
KT021567, KT316760, KT594680, KY992442; N. cyclamineus DC., N. cyclamineus 1, Portugal, S. Joao do Monte, Marques 1753 & Draper, leaf sample provided
by C. Silva, KT340523, KT021573, KT316766, KT594686, KY992432; N. cyclamineus 2, Spain, Santiago de Compostela, Marques 1790 & Draper, leaf sample
provided by M. Serrano, KT340525, KT021575, KT316768, KT594688, –; N. cyclamineus 3, Spain, Val do Dubra, Ínsua, M. Serrano s.n., leaf sample provided by
M. Serrano, KT340524, KT021574, KT316767, KT594687, –; N. ×hannibalis A.Fern., N. ×hannibalis (hispanicus × triandrus subsp. lusitanicus), Portugal, Águas
Belas, pers. coll. A. Fernandes, KT340545, KT021599, KT316792, KT594712, KY992417; N. hispanicus Gouan, N. hispanicus, Portugal, Águas Belas, Marques
1950 & Draper (LISU 181999), KT340544, KT021598, KT316791, KT594711, KY992416; N. hispanicus subsp. perez-chiscanoi (Fern, Casas) Fern, Casas, N. his-
panicus subsp. perez-chiscanoi, BCN920175, germplasm coll. of BCN, KT799074, KT033086, KT808819, KT878425, KY992415; N. jacetanus Fern.Casas,
N. jacetanus, Spain, San Juan de la Peña, leaf sample provided by L. Villar, KT340546, KT021600, KT316793, KT594713, KY992426; N. longispathus Pugsley,
N. longispathus 1, Spain, Arroyo de la Venta, Medrano & Herrera s.n., leaf sample provided by C. Herrera, KT340550, KT021613, KT619086, KT594726, KY992400;
N. longispathus 2, Spain, Sierra de Cazorla, Fuente del Oso, Medrano & Herrera 2005-06, leaf sample provided by C. Herrera, KT340549, KT021612, KT619085,
KT594725, KY992402; N. longispathus 3, Spain, Sierra de Cazorla, La Cabrilla, Medrano & Herrera 2006-10, leaf sample provided by C. Herrera, KT340551,
KT021614, KT619087, KT594727, KY992403; N. longispathus 4, Spain, Sierra de Cazorla, Valdecuevas, Medrano & Herrera 2006-09, leaf sample provided by
C. Herrera, KT340552, KT021615, KT619088, KT594728, –; N. longispathus 5, Spain, Villaverde de Guadalimar, Medrano & Herrera 2006-02, leaf sample provided
by C. Herrera, KT340553, KT021616, KT619089, KT594729, –; N. macrolobus (Jord.) Pugsley, N. macrolobus, BCN930825, germplasm coll. of BCN, KT799113,
KT021619, KT619092, KT594732, KY992411; N. minor L., N. minor, Spain, Escurial, Pico Cervero, A. González de Castro s.n., leaf sample provided by A. González
de Castro, KT799114, KT021620, KT619093, KT594733, KY992427; N. moleroi Fern.Casas, N. moleroi, Spain, Nuria, germplasm coll. of BCN, KT210959, KR872977,
KT272909, KT279965, KY992441; N. nanus (Haw.) Spach, N. nanus, living coll. of JB-MNHN s.n., KT799116, KT021622, KT619095, KT594735, KY992428;
N. nevadensis Pugsley, N. nevadensis 1, Spain, Haza de las Piedras, J.B. Hoya de Pedraza s.n., J.B. Hoya de Pedraza, KT799122, KT021628, KT619101, KT594741,
KY992406; N. nevadensis 2, Spain, Prados del Aire, J.B. Hoya de Pedraza s.n., J.B. Hoya de Pedraza, KT799123, KT021629, KT619102, KT594742, KY992407;
N. nevadensis 3, Spain, Barranco del Primer Roble, J.B. Hoya de Pedraza s.n., J.B. Hoya de Pedraza, KT799124, KT021630, KT619103, KT594743, KY992408;
N. nevadensis 4, Spain, Sierra Nevada, J.B. Hoya de Pedraza s.n., J.B. Hoya de Pedraza, KT799127, KT021633, KT619106, KT594746, KY992409; N. nevadensis 5,
Spain, Barranco del Agudero, Medrano & Herrera s.n., leaf sample provided by C.  Herrera, KT799125, KT021631, KT619104, KT594744, KY992410;

Version of Record 853

Marques & al. • A three-genome phylogeny of Narcissus (Amaryllidaceae) TAXON 66 (4) • August 2017: 832–854

Appendix 1. Continued.
N. nevadensis 6, Spain, Barranco de los Tejos, J.B. Hoya de Pedraza s.n., J.B. Hoya de Pedraza, KT799121, KT021627, KT619100, KT594740, –; N. nevadensis 7,
Spain, Cortijo del Sotillo, Medrano & Herrera s.n., leaf sample provided by C. Herrera, KT799126, KT021632, KT619105, KT594745, –; N. nobilis (Haw.) Schult.
& Schult.f., N. nobilis 1, Spain, Puerto del Pontón, H. Nava & A. Bueno s.n., leaf sample provided by H. Nava, KT799117, KT021623, KT619096, KT594736, KY992436;
N. nobilis 2, Spain, Puerto de Tarna, H. Nava & A. Bueno s.n., leaf sample provided by H. Nava, KT799118, KT021624, KT619097, KT594737, –; N. nobilis var.
leonensis (Pugsley) A.Fern., N. nobilis var. leonensis 1, Spain, Puerto de Lunada, H. Nava & A. Bueno s.n., leaf sample provided by H. Nava, KT340548, KT021611,
KT619084, KT594724, KY992435; N. nobilis var. leonensis 2, Spain, Cerneja, H. Nava & A. Bueno s.n., leaf sample provided by H. Nava, KT340547, KT021610,
KT619083, KT594723, –; N. pallidiflorus Pugsley, N. pallidiflorus 1, Spain, Villarreal de Alava, H. Nava & A. Bueno s.n., leaf sample provided by H. Nava, KT799120,
KT021626, KT619099, KT594739, KY992430; N. pallidiflorus 2, Spain, Covadonga, living coll. of RJB s.n., KT799119, KT021625, KT619098, KT594738, –; N. por­
tensis Pugsley, N. portensis, BCN920206, germplasm coll. of BCN, KT799135, KT033074, KT808807, KT878413, KY992413; N. primigenius (Fern.Suárez ex
Laínz) Fern.Casas & Laínz, N. primigenius 1, Spain, Gijón, H. Nava & A. Bueno s.n., leaf sample provided by H. Nava, KT799136, KT033075, KT808808, KT878414,
KY992434; N. primigenius 2, Spain, Covadonga, H. Nava & A. Bueno s.n., leaf sample provided by H. Nava, KT799137, KT033076, KT808809, KT878415, –;
N. pseudonarcissus L., N. pseudonarcissus 1, Andorra, J. Pedrol 7951, pers. coll. J. Pedrol, KT799066, KT033077, KT808810, KT878416, KY992418; N. pseudonar-
cissus 2, Belgica, Sibret, Belleau, C. Lefèbvre s.n., leaf sample provided by C. Lefèbvre, KT799067, KT033078, KT808811, KT878417, KY992419; N. pseudonarcis-
sus 3, Belgium, Grand Leez, C. Lefèbvre s.n., leaf sample provided by C. Lefèbvre, KT799068, KT033079, KT808812, KT878418, –; N. pseudonarcissus 4, Belgium,
Onoz, C. Lefèbvre s.n., leaf sample provided by C. Lefèbvre, KT799069, KT033080, KT808813, KT878419, –; N. pseudonarcissus 5, Spain, Soria, F.L. Domínguez
s.n., leaf sample provided by F.L. Domínguez, KT799070, KT033081, KT808814, KT878420, –; N. pseudonarcissus 6, Spain, Sierra de Urbasa, J.C. Moreno s.n.,
leaf sample provided by J.C. Moreno, KT799071, KT033082, KT808815, KT878421, –; N. pseudonarcissus subsp. eugeniae (Fern.Casas) Fern.Casas, N. pseudo-
narcissus subsp. eugeniae 1, Spain, Sierra de Cucalón, Fonfria, leaf sample provided by M. Serrano, KT340528, KT021582, KT316775, KT594695, KY992433;
N. pseudonarcissus subsp. eugeniae 2, Spain, Sierra de Moncayo, living coll. of JB-MNHN, KT340526, KT021580, KT316773, KT594693, –; N. pseudonarcissus
subsp. eugeniae 3, Spain, Puerto de las Señales, H. Nava & A. Bueno s.n., leaf sample provided by H. Nava, KT340527, KT021581, KT316774, KT594694, –;
N. pseudonarcissus subsp. munozii-garmendiae (Fern.Casas) Fern.Casas, N. pseudonarcissus subsp. munozii-garmendiae, BCN920185, germplasm coll. of BCN,
KT799115, KT021621, KT619094, KT594734, KY992414; N. pumilus Salisb., N. pumilus, Portugal, living coll. of JB-MNHN s.n., KT326980, KT210966, KT279996,
KT340505, KY992429; N. segurensis Ríos, D.Rivera, Alcaraz & Obón, N. segurensis, Spain, Arroyo del Tejuelo, Medrano & Herrera 2005-01, leaf sample provided
by C. Herrera, KT799085, KT033097, KT808830, KT878436, KY992396/​KY992397; N. ×somedanus Fern.Casado & al., N. ×somedanus (astu­riensis × triandrus),
Spain, Somiedo, M. Serrano s.n., leaf sample provided by M. Serrano, KT340543, KT021597, KT316790, KT594710, KY992425; N. tortuosus (Haw.) Spach, N. tor-
tuosus, Spain, Cabo Peñas, H. Nava s.n., pers. coll. H. Nava, KT863431, KT033127, KT808860, KT878466, KY992431; N. yepesii Ríos & al., N. yepesii 1, Spain,
Lower Arroyo de Navalasna, Medrano & Herrera 2005-02, leaf sample provided by C. Herrera, KT863445, KT033141, KT808874, KT878480, KY992398; N. yepesii 2,
Spain, Upper Arroyo de Navalasna, Medrano & Herrera 2005-03, leaf sample provided by C. Herrera, KT863446, KT033142, KT808875, KT878481, KY992399;
N. yepesii 3, Spain, Fuente de la Jordana, Medrano & Herrera 2005-04, leaf sample provided by C. Herrera, KT863447, KT033143, KT808876, KT878482, –;
N. yepesii 4, Spain, Navalcaballo, Medrano & Herrera 2005-05, leaf sample provided by C. Herrera, KT863448, KT033144, KT808877, KT878483, –; — Section
Serotini: N. serotinus L., N. serotinus 1, Portugal, Ponte de Ajuda, Marques 1156 & Rosseló-Graell (LISU 187994), HM371975.1, KM349545.1, HM231911.1,
HM174804.1, KY971379; N. serotinus 2, Portugal, Montes Juntos, Marques 1018 (LISU 188000), HM372005.1, KM349547.1, HM231941.1, HM174834.1, KY971380;
N. serotinus 3, Portugal, Póvoa de Sao Miguel, Marques 2002 & Draper (LISU 187629), HM372065.1, KM349551.1, HM232001.1, HM174894.1, KY971381; N. sero-
tinus 4, Portugal, Vidigueira, Marques 1013 (LISU 173710), HM372110.1, KM349554.1, HM232046.1, HM174939.1, KY971382; — Section Tazettae: N. dubius
Gouan, N. dubius 1, Spain, Monegros. MIM95020043, living coll. of MIM, KT124413, KT021577, KT316770, KT594690, KY971396; N. dubius 2, Spain, Coll Baix.
BCN s.n., germplasm coll. of BCN, KT124414, KT021578, KT316771, KT594691, KY971397; N. dubius 3, Spain, Caravaca de la Cruz, living coll. of RJB s.n.,
KT124412, KT021576, KT316769, KT594689, –; N. dubius 4, Spain, Grazalema, S.L. Jury 10417a. RNG s.n., leaf sample provided by S.L. Jury, KT124415, KT021579,
KT316772, KT594692, –; N. elegans (Haw.) Spach, N. elegans 1, Morocco, Ksar-el-Kebir, living coll. of JB-MNHN 1200, HM372679.1, KM349596.1, HM232615.1,
HM175508.1, KY971383; N. elegans 2, Morocco, Lalla Minouna, living coll. of JB-MNHN 1201, HM372709.1, KM349598.1, HM232645.1, HM175538.1, KY971384;
N. elegans 3, Morocco, Moulay-Idriss, living coll. of JB-MNHN 1196, HM372739.1, KM349600.1, HM232675.1, HM175568.1, KY971385; N. elegans 4, Morocco,
Cap Martil, living coll. of JB-MNHN 1193, HM372724.1, KM349599.1, HM232660.1, HM175553.1, KY971386; N. elegans 5, Italy, Sicily, S.L. Jury 10252. RNG
s.n, leaf sample provided by S.L. Jury, HM372859.1, KM349608.1, HM232795.1, HM175688.1, –; N. elegans 6, Morocco, Taza, pers. coll. A. Mohammed, HM372799.1,
KM349604.1, HM232735.1, HM175628.1, –; N. miniatus Donn.-Morg. & al., N. miniatus 1, Spain, Medina Sidonia. (LISU 187637), HM372125.1, KM349555.1,
HM232061.1, HM174954.1, KY971387; N. miniatus 2, Spain, Los Belones, living coll. of JB-MNHN 1146, HM372275.1, KM349567.1, HM232211.1, HM175104.1,
KY971388; N. miniatus 3, Spain, La Línea de la Concepcion, living coll. of JB-MNHN 1138, HM372260.1, KM349566.1, HM232196.1, HM175089.1, KY971389;
N. miniatus 4, Spain, Majorca, living coll. of JBS s.n., HM372456.1, KM349580.1, HM232392.1, HM175285.1, KY971390; N. pachybolbus Durieu, N. pachybolbus,
Spain, Taza, S.L. Jury 15748. RNG s.n., leaf sample provided by S.L. Jury, KT272885, KT033056, KT808789, KT594751, KY930348; N. pannizianus Parl., N. pan-
nizianus 1, Spain. BCN890151, germplasm coll. of BCN, KT272887, KT033058, KT808791, KT594753, KY930350; N. pannizianus 2, Spain, Puerto de Encinas
Borrachas. BCN s.n., germplasm coll. of BCN, KT272886, KT033057, KT808790, KT594752, –; N. papyraceus Ker Gawl., N. papyraceus 1, Portugal, Azambuja,
Marques 1070 & Draper (LISU 188018), KT272893, KT033064, KT808797, KT594759, KY930344; N. papyraceus 2, Portugal, Vila do Bispo, Marques 800 &
Draper (LISU 188028), KT272888, KT033059, KT808792, KT594754, KY930346; N. papyraceus 3, Morocco, Larache, S.L. Jury 16231. RNG s.n., leaf sample
provided by S.L. Jury, KT272890, KT033061, KT808794, KT594756, KY930345; N. papyraceus 4, Portugal, Fonte Benémola, Marques 50, Brehm & Draper (LISU
188038), KT272891, KT033062, KT808795, KT594757, KY930347; N. papyraceus 5, Portugal, Ribeira de Quarteira, Marques 1061 & Draper (LISU 188082),
KT272892, KT033063, KT808796, KT594758, –; N. papyraceus 6, Spain, Algibre, Marques 2045 (LISU 188038), KT272889, KT033060, KT808793, KT594755, –;
N. papyraceus 7, Portugal, Algibre, Marques 1065 & Draper, LSIU 188030, KT272894, KT033065, KT808798, KT594760, –; N. papyraceus 8, Morocco, Temara,
pers. coll. A. Mohammed, KT272895, KT033066, KT808799, KT594761, –; N. papyraceus 9, Morocco, Tanger, pers. coll. A. Mohammed, KT272896, KT033067,
KT808800, KT594762, –; N. papyraceus 10, Spain, Carcabuey, E. Triano s.n., pers. coll. E. Triano, KT272897, KT033068, KT808801, KT594763, –; N. polyanthos
Loisel., N. polyanthos, Morocco, Moulay-Idriss, S.L. Jury 13714, RNG s.n., leaf sample provided by S.L. Jury, KT799134, KT033073, KT808806, KT594768,
KY930349; N. ×pujolii Font Quer, N. ×pujolii (assoanus × dubius), Spain, Coll Baix, Marques 2222 & Draper, germplasm coll. of BCN, KT799133, KT033070,
KT808803, KT594765, KY971395/​KY971422; N. tazetta L., N. tazetta 1, Portugal, Dornes, Águas Belas, Marques 1999 & Draper (LISU 265714), KT799087,
KT033099, KT808832, KT878438, KY930339; N. tazetta 2, Portugal, Santa Iria, Santarem, Marques 2000 & Draper (LISU 265715), KT799088, KT033100,
KT808833, KT878439, KY930340; N. tazetta 3, Japan. BCN890576, germplasm coll. of BCN, KT799091, KT033103, KT808836, KT878442, KY930341; N. tazetta 4,
Spain, Majorca, living coll. of JBS, KT799089, KT033101, KT808834, KT878440, KY930342; N. tazetta 5, Spain, Teulada, J. Soler & D. Draper s.n., pers. coll.
J. Soler, KT799090, KT033102, KT808835, KT878441, KY930343; N. tortifolius Fern.Casas, N. tortifolius 1, Spain. BCN890565, germplasm coll. of BCN, KT863427,
KT033123, KT808856, KT878462, KY971392/​KY971394; N. tortifolius 2, Spain. MIM s.n., living coll. of MIM, KT863428, KT033124, KT808857, KT878463,
KY971393; N. tortifolius 3, Spain, Los Castaños, J.F. Mota s.n., leaf sample provided by J.F. Mota, KT863429, KT033125, KT808858, KT878464, –; N. tortifolius 4,
Spain, Sorbas, J.F. Mota s.n., leaf sample provided by J.F. Mota, KT863430, KT033126, KT808859, KT878465, –; — Outgroups: Asparagus cochinchinensis
(Lour.) Merr., Asparagus cochinchinensis, used only to date the phylogeny, JX903371.1, AB029804.1, –, –, –; Lapiedra martinezii Lag., Lapiedra martinezii, Spain,
Gata de Gorgos, pers. coll. J. Soler, KM359734, KM359730, KT279981, KT279950, KY971418; Leucojum aestivum L., Leucojum aestivum, living coll. of RJB s.n.,
KM359736, KM359732, KT279983, KT279952, KY971419; Nerine bowdenii Watson, Nerine bowdenii, living coll. of RJB s.n., KM359737, KM359733, KT279984,
KT279953, KY971421; Nothoscordum dialy­stemon (Guagl.) Crosa, Nothoscordum dialystemon, used only to date the phylogeny, AF508406.1, JQ435522.1, –, –, –;
Sternbergia lutea (L.) Ker-Gawl. ex Spreng., Sternbergia lutea, living coll. of RJB s.n., KM359735, KM359731, KT279982, KT279951, KY971420.

854 Version of Record