ArticlePDF Available

Phylogeny of the American Amaryllidaceae Based on nrDNA ITS Sequences

  • Arizona State University (ASU) and Montgomery Botanical Center (MBC)

Abstract and Figures

Analysis of three plastid DNA sequences for a broad sampling of Amaryllidaceae resolve the American genera of the Amaryllidaceae as a clade that is sister to the Eurasian genera of the family, but base substitution rates for these genes are too low to resolve much of the intergeneric relationships within the American clade. We obtained ITS rDNA sequences for 76 species of American Amaryllidaceae and analyzed the aligned matrix cladistically, both with and without gaps included, using two species of Pancratium as outgroup taxa. ITS resolves two moderately to strongly supported groups, an Andean tetraploid clade, and a primarily extra-Andean “hippeastroid” clade. Within the hippeastroid clade, the tribe Griffineae is resolved as sister to the rest of Hippeastreae. The genera Rhodophiala and Zephyranthes are resolved as polyphyletic, but the possibility of reticulation within this clade argues against any re-arrangement of these genera without further investigation. Within the Andean subclade, Eustephieae resolves as sister to all other tribes; a distinct petiolate-leafed group is resolved, combining the tribe Eucharideae and the petiolate Stenomesseae; and a distinct Hymenocallideae is supported. These Andean clades are all at least partially supported by plastid sequence data as well. We infer from our data that a great deal of the diversity of the family in the Americas is recent, and that the American Amaryllidaceae may have been reduced to peripheral isolates some time after its initial entry and spread through the Americas. While the sister relationship of the American and Eurasian clades might argue for a Boreotropical origin for the family in America, the cladistic relationships within the American clade based on ITS do not provide any further support for this or any other hypothesis of the family's entry into America. The new tribe Clinantheae is described (four genera: Clinanthus, Pamianthe, Paramongaia, and Pucara), and the lorate-leafed species of Stenomesson are transferred to Clinanthus. Communicating Editor: Kathleen A. Kron
Content may be subject to copyright.
Systematic Botany (2000), 25(4): pp. 708–726
qCopyright 2000 by the American Society of Plant Taxonomists
Phylogeny of the American Amaryllidaceae based on
nrDNA ITS sequences
W. M
USDA-ARS-SHRS, 13601 Old Cutler Road, Miami, FL 33158 and Fairchild Tropical Garden,
10901 Old Cutler Road, Miami, Florida 33158
L. G
and Q
University of Florida, Department of Environmental Horticulture, 1545 Fifield Hall,
Gainesville, Florida 32611
University of Florida, Fort Lauderdale Research and Education Center, 3205 College Avenue,
Fort Lauderdale, Florida 33314
Communicating Editor: Kathleen A. Kron
.Analysis of three plastid DNA sequences for a broad sampling of Amaryllidaceae resolve the
American genera of the Amaryllidaceae as a clade that is sister to the Eurasian genera of the family, but base
substitution rates for these genes are too low to resolve much of the intergeneric relationships within the
American clade. We obtained ITS rDNA sequences for 76 species of American Amaryllidaceae and analyzed
the aligned matrix cladistically, both with and without gaps included, using two species of Pancratium as
outgroup taxa. ITS resolves two moderately to strongly supported groups, an Andean tetraploid clade, and
a primarily extra-Andean ‘‘hippeastroid’’ clade. Within the hippeastroid clade, the tribe Griffineae is resolved
as sister to the rest of Hippeastreae. The genera Rhodophiala and Zephyranthes are resolved as polyphyletic,
but the possibility of reticulation within this clade argues against any re-arrangement of these genera without
further investigation. Within the Andean subclade, Eustephieae resolves as sister to all other tribes; a distinct
petiolate-leafed group is resolved, combining the tribe Eucharideae and the petiolate Stenomesseae; and a
distinct Hymenocallideae is supported. These Andean clades are all at least partially supported by plastid
sequence data as well. We infer from our data that a great deal of the diversity of the family in the Americas
is recent, and that the American Amaryllidaceae may have been reduced to peripheral isolates some time
after its initial entry and spread through the Americas. While the sister relationship of the American and
Eurasian clades might argue for a Boreotropical origin for the family in America, the cladistic relationships
within the American clade based on ITS do not provide any further support for this or any other hypothesis
of the family’s entry into America. The new tribe Clinantheae is described (four genera: Clinanthus, Pamianthe,
Paramongaia, and Pucara), and the lorate-leafed species of Stenomesson are transferred to Clinanthus.
Amaryllidaceae J. St.-Hil., a cosmopolitan (pre-
dominantly pantropical) family of petaloid mono-
cots, represent one of the elements of the Linnaean
Hexandria monogynia (Linnaeus 1753), the 51 genera
of which have been variously classified since as lil-
iaceous or amaryllidaceous. Meerow et al.(1999) re-
viewed the taxonomic history of the family. Despite
a lack of consensus on generic limits and tribal de-
limitations within Amaryllidaceae, cladistic analy-
sis has only rarely been applied to problems in the
family, such as by Nordal and Duncan (1984) for
Haemanthus and Scadoxus, two closely related, bac-
cate-fruited African genera; Meerow (1987, 1989)
for Eucrosia and Eucharis and Caliphruria respective-
ly; Snijman (1994) and Snijman and Linder (1996)
for various taxa of tribe Amaryllideae; and Meerow
et al. (1999) for the entire family using three plastid
DNA sequences. Phylogenetic studies for the entire
family using morphological characters are made
difficult by homoplasy for many conspicuous char-
acters within this highly canalized group (Meerow
et al., 2000).
The four most recent infrafamilial classifications
of Amaryllidaceae are those of Traub (1963), Dahl-
gren et al. (1985), Mu¨ ller-Doblie s and Mu¨ ller-Dob-
lies (1996) and Meerow and Snijman (1998). Traub’s
scheme included Alliaceae, Hemerocallidaceae, and
Ixioliriaceae as subfamilies, following Hutchinson
(1934, 1959) in part. Within his subfamily Amar-
ylloideae, he erected two informal taxa, ‘‘infrafam-
ilies’’ Amarylloidinae and Pancratioidinae, both of
which were polyphyletic (Meerow 1995). Dahlgren
et al. (1985) dispensed with any subfamilial classi-
fication above the level of tribe, recognizing eight,
and treated as Amaryllidaceae only those genera in
Traub’s Amarylloideae. Stenomesseae and Euste-
phieae were combined. Meerow (1995) resurrected
Eustephieae from Stenomesseae and suggested that
two new tribes needed to be recognized, Calostem-
mateae and Hymenocallideae. Mu¨ ller- Doblies and
Mu¨ ller-Doblies (1996) recognized ten tribes (among
them Calostemmateae) and nineteen subtribes,
many of them monogeneric; Meerow and Snijman
(1998) recognized 14 tribes, with two subtribes only
in one of them. These classifications are compared
in detail in Meerow et al. (1999).
Meerow et al. (1999) presented cladistic analyses
of combined plastid DNA sequences rbcL, trnL gene
and the trnL-F intergenic spacer for 48 genera of
Amaryllidaceae and 29 genera of related Aspara-
gales. Good support was provided for the mono-
phyly of Amaryllidaceae. The infra-familial rela-
tionships of Amaryllidaceae were resolved along
biogeographic lines. Ito et al. (1999) resolved a very
similar topology for a more limited sampling of
Amaryllidaceae and related asparagoids using
plastid matK sequences. The most surprising con-
clusion from the plastid sequence phylogeny was
that the Eurasian and American elements of the
family are each monophyletic sister clades. How-
ever, there was insufficient base substitution among
the American genera to resolve their relationships
sufficiently with plastid sequences.
The 18S and 26S subunits of nuclear ribosomal
DNA (nrDNA) are separated by two internal tran-
scribed spacer regions, ITS1 and ITS2, and a 5.8S
gene located between the two ITS regions. ITS1
varies in angiosperms from 187–298 bp, and ITS2
from 187–252 bp (Baldwin et al. 1995). The two ITS
regions evolve more rapidly than the coding re-
gions they separate. For closely related taxa, rapidly
evolving, non-coding regions of nuclear genes such
as ITS can potentially yield a greater degree of in-
formative sequence variation than the more highly
conserved coding regions. From a practical stand-
point, the small size of the ITS region, and its lo-
cation between highly conserved sequences, make
the spacers easy to amplify by PCR (Baldwin et al.
1995). Potential problems of paralogous sequences
due to the multiple copy nature of nrITS can be at
least partially overcome by sequencing more than
one clone of each taxon or pooling PCR extracts of
several clones for the sequence reactions (Baldwin
et al. 1995).
ITS shows greatest utility for generating gene
phylogenies at the rank of family and below (Bald-
win et al. 1995). To date, ITS sequences have been
sparingly successful in resolving family-wide phy-
logenies (Fouquieriaceae, Schultheis and Baldwin
1999; Nothofagaceae, Manos 1993; Winteraceae, Suh
et al. 1992), but have been most valuable when ap-
plied to single genera or at the subfamilial rank;
e.g., Aquilegia (Ro et al. 1997), Fraxinus (Jeandroz et
al. 1997), Lupinus (Kass and Wink 1997), Saintpaulia
(Moeller and Cronk 1997), Adoxaceae: Adoxoideae
(Eriksson and Donoghue 1997), Rosaceae subfamily
Maloideae (Campbell et al. 1995), Orchidaceae sub-
tribe Orchidinae (Bateman et al. 1997; Pridgeon et
al. 1997), Poaceae subfamily Arundinoideae (Hsiao
et al. 1998).
In this paper we present cladistic analyses of the
internal transcribed spacer region (ITS) of nuclear
ribosomal DNA for 76 species of American Amar-
yllidaceae (representing all but five of the American
endemic genera) and discuss our results in the con-
text of previous plastid sequence phylogenies (Ito
et al. 1999; Meerow et al. 1999). We also discuss the
biogeographic implications of these data.
Plant Materials. Species used in the sequence
analyses, voucher specimens and GenBank acces-
sion numbers for the sequences are listed in Table
1. All genera of the American Amaryllidaceae were
sampled except Mathieua (presumed extinct), Pla-
giolirion (repeated attempts to amplify DNA from a
recent herbarium specimen failed) and Placea, Pu-
cara, and Traubia (material not available). The pres-
ence of the pantropical Crinum (tribe Amaryllideae)
in the Americas is understood to represent a dis-
persal event unrelated to the origins of the endemic
American tribes of Amaryllidaceae (Arroyo and
Cutler 1984; Meerow et al. 1999) and no American
species of this genus are considered here.
DNA Extraction. Genomic DNA was extracted
from silica gel dried leaf tissue using a modified
CTAB procedure of Doyle and Doyle (1987).
PCR and Sequencing Protocol. Amplification of
the ribosomal DNA ITS1/5.8S/ITS2 region was ac-
complished using flanking primers (18S, 26S)
AB101 and AB102 (Douzery et al. 1999), and the
original White et al. (1990) primers ITS5, 4, 2 and
3 to amplify the spacers along with the intervening
5.8S sequence. Amplified products were purified
1. Taxa used in the cladistic analyses nrDNA (ITS) sequence. Tribal and generic assignments in Amarylli-
daceae follow Meerow and Snijman (1998). All vouchers are deposited at FTG unless otherwise noted.
Taxon Voucher GenBank
Accession No.
Caliphruria korsakoffii (Traub) Meerow Meerow 1096 (FLAS) AF223529
Caliphruria subedentata Bak. Meerow 1109 (FLAS) AF223549
Eucharis amazonica Lind. ex Planch. Schunke 14057 (FLAS) AF223538
Eucharis castelnaeana (Baill) Macbr. Schunke 14156 (FLAS) AF223525
Eucharis formosa Meerow Whitten et al. 95020 (FLAS) AF223539
Urceolina microcrater Kra¨nzl. Plowman & Kennedy 5721 (GH) AF223531
Chlidanthus boliviensis Traub & I. S. Nelson Traub 529a (MO) AF223545
Chlidanthus fragrans Herb. Meerow 2312 AF223524
Eustephia darwinii Vargas Meerow 2436 AF223543
Hieronymiella argentina (Pax) A. T. Hunz. & S.
M. W. Chase 1901 (K) AF223541
Griffinia hyacinthina Ker Gawler Meerow 2106 AF223473
Griffinia noctrurna Rav. Pereira & Paula 2351 (UB) AF223491
Griffinia rochae Morel Meerow 1154 AF223486
Habranthus brachyandrus (Bak.) Sealy Meerow 2400 AF223504
Habranthus immaculatus Traub & Clint Meerow 2401 AF223500
Habranthus martinezii Rav. Meerow 2437 AF223497
Habranthus sp. Meerow 2402 AF223499
Habranthus tubispathus (L’He´ rit.) Traub Meerow 2403 AF223498
Haylockia americana (Hoffgg.) Herter M. W. Chase 3585 (K) AF223506
Hippeastrum blumenavium (K. Koch & Bouche ex
Carr) Sealy
Meerow 2404 AF223501
Hippeastrum brasilianum (Traub & Doran) Dutilh Meerow 2405 AF223479
Hippeastrum macbridei (Vargas) Gereau & Brako Meerow 2435 AF223509
Hippeastrum molleviquensis (Ca´ rd.) Van Scheepen Doran 1538 (MO) AF223489
Hippeastrum papilio (Rav.) Van Scheepen Meerow 2406 AF223496
Hippeastrum parodii A. T. Hunz. & Cocucci Meerow 2434 AF223508
Hippeastrum reticulatum Herb. Meerow 2407 AF223484
Phycella ignea Lindl. Meerow 2408 AF223505
Pyrolirion sp. M. W. Chase 3639 (K) AF223493
Rhodophiala bagnoldii (Herb.) Traub Meerow 2425 AF223476
Rhodophiala bifida (Herb.) Traub Meerow 2410 AF223477
Rhodophiala chilensis (L’He´rit.) Traub Meerow 2426 AF223480
Rhodophiala moelleri (R. Phil.) Traub M. W. Chase 1908 (K) AF223481
Sprekelia formosissima (L.) Herb. Meerow 1151 AF223483
Worsleya rayneri (Hook.) Traub & Moldenke Meerow 2411 AF223475
Zephyranthes atamasco (L.) Herb. Meerow 2412 AF223474
Zephyranthes simpsonii Meerow 2413 AF223472
Zephyranthes candida Herb. Meerow 2414 AF223503
Zephyranthes cearensis Bak. Meerow 2415 AF223507
Zephyranthes citrina Bak Meerow 2416 AF223495
Zephyranthes drummondii D. Don. Meerow 2417 AF223488
Zephyranthes filifolia Herb. ex Baker M. W. Chase 1836 (K) AF223490
Zephyranthes flavissima Rav. Meerow 2418 AF223510
Zephyranthes grandiflora Lindl. Meerow 2419 AF223485
Zephyranthes mesochloa Herb. Meerow 2420 AF223502
Zephyranthes morrisclintii Traub & T. M. Howard Meerow 2421 AF223492
Zephyranthes pulchella F. D. Smith Meerow 2422 AF223494
Zephyranthes rosea Lindl. Meerow 2429 AF223487
Zephyranthes smallii (Alexander) Traub Meerow 2423 AF223482
1. Continued.
Taxon Voucher GenBank
Accession No.
Hymenocallis acutifolia (Herb.) Sweet Meerow 2424 AF223514
Hymenocallis glauca M. Roem. Meerow 2433 AF223515
Hymenocallis latifolia (Mill.) Roem. Meerow 2438 AF223516
Hymenocallis speciosa Salisb. Meerow 2439 AF223512
Hymenocallis tubiflora Salisb Meerow 2440 AF223513
Ismene hawkesii (Vargas) Gereau & Meerow Meerow 2441 AF223519
Ismene longipetala (Lindl.) Meerow Sagastegui 15454 AF223520
Ismene narcissiflora Jacq. Meerow 2306 AF223518
Ismene vargasii (Velarde) Gereau & Meerow Meerow 2308 AF223517
Leptochiton quitoensis (Herb.) Sealy Meerow 1116 AF223521
Pancratium canariensis L. Meerow 1142 AF223531
Pancratium tenuifolium Hochst. Meerow 2427 AF223537
Eucrosia bicolor Ker Gaw l. Meerow 1113 AF223511
Eucrosia dodsonii Meerow & Dehgan Meerow 1115 AF223544
Eucrosia eucrosioides (Pax) Traub Meerow 1117 AF223548
Eucrosia stricklandii (Bak.) Meerow Meerow 2428 AF223540
Pamianthe peruviana Stapf Meerow 2304 AF223546
Paramongaia weberbaueri Velarde Meerow 2303 AF223536
Phaedranassa cinerea Rav. Meerow & Meerow 1045 AF223527
Phaedranassa tunguraguae Rav. Meerow & Meerow 1130 AF223526
Phaedranassa ventricosa Bak. Meerow 1129 AF223528
Rauhia decora Ravenna Meerow 1160 AF223523
Rauhia multiflora (Kunth) Rav. Meerow 2441 AF223522
Stenomesson aurantiacum Herb. Meerow 1061 (FLAS) AF223542
Stenomesson flavum (R. & P.) Herb. Meerow 2430 AF223532
Stenomesson humilis Bak. Meerow 2442 AF223547
Stenomesson incarnatum Bak. Meerow 1120 AF223533
Stenomesson mirabile Rav. S. Leiva et al. 2000 (HAO) AF223534
Stenomesson variegatum (R. & P.) Bak. Meerow 1159 AF223535
using QIAquick (Qiagen, Valencia, CA) columns,
following manufacturers protocols. PCR amplifica-
tions were performed on an ABI 9700 (Perkin-Elmer
Applied Biosystems, Inc., Foster City, CA), running
28 cycles of the following program: 4 min at 94
1 min at 52
, and 3 min at 72
Cycle sequencing reactions were performed di-
rectly on purified PCR products in the ABI 9700,
using standard dideoxy cycle protocols for sequenc-
ing with dye terminators on an ABI 377 automated
sequencer (according to the manufacturer’s proto-
cols; Applied Biosystems, Inc.). It was determined
via cloning prior to beginning our ITS work on the
neotropical Amaryllidaceae that paralogous varia-
tion among the tandem repeats of the ITS region
(Baldwin et al. 1995) is not an issue in the group.
Sequence Alignment. ITS sequences were aligned
in two ways. First, we used the sequence editing and
alignment program Sequencher (Gene Codes, Inc.,
Ann Arbor, MI) to align sequences of closely related
taxa with subsequent builds of these smaller align-
ments performed manually. We also used the pro-
gram ClustalX (Higgins and Sharp 1988, Thompson
et al. 1997) to align the sequences with a gap open-
ing penalties of both 15 and 30 and gap extension
penalties of 0.5 to 6.66. We also loaded our Sequencher
alignment into ClustalX, and allowed the program to
re-align the sequences but without removing gaps we
had previously inserted. Low-scoring areas in all of
the alignments were carefully examined. By compar-
ing alignments among the three methods, as well as
running heuristic searches with PAUP (see below) on
each reiteration, we were able to refine our ultimate
alignment considerably before cladistic analysis. Our
manually edited alignment consistently produced the
shortest trees when analyzed cladistically. The matrix
and resulting trees are available from TreeBASE (http:
Analyses. Aligned matrices were analyzed us-
ing the parsimony algorithm of the software pack-
age PAUP* for Macintosh (v4.02 beta; Swofford
1998), with the MULPARS option and ACCTRAN
optimization invoked. Zero length branches were
collapsed only if the maximum length 50. A suc-
cessive weighting (SW) strategy (Farris 1969) was
implemented. SW is a useful tool employed to glob-
ally reduce the effect of highly homoplasious base
positions on the resulting topologies (Lledo´etal.
1998; Meerow et al. 1999; Wenzel 1997). Whole cat-
egory weights (codon or transversion) exhibit broad
and overlapping ranges of consistency (Olmstead
1997), whereas successive weighting independently
assesses each base position of the multiple align-
ment based on their consistency in the initial anal-
The initial tree search was conducted under the
Fitch (equal) weights (Fitch 1971) criterion with
1000 random sequence additions and SPR (subtree
pruning-regrafting) branch swapping, permitting
only ten trees to be held at each step to reduce the
time spent searching trees at suboptimal levels. All
trees collected in the 1000 replicates were swapped
to completion or an upper limit of 5000 trees. The
characters were then reweighted by the rescaled
consistency index, and a further 50 replications of
random sequence additions conducted with the
weighted matrix saving 15 trees per replication.
These trees were then swapped to completion or an
upper limit of 5000 trees. The resulting trees were
then used to reweight the matrix a second time by
the rescaled consistency index, and another 50 rep-
licates of random sequence addition conducted,
saving 15 trees per replication, with subsequent
swapping on those trees. This cycle was repeated
until two successive rounds found trees of the same
Internal support was determined by bootstrap-
ping (5000 replicates) with the final reweighted
character matrix, and with 5000 replications of the
jackknife procedure (Farris et al. 1996), the later
with equal weight applied. In this way, support for
both the weighted and unweighted matrices could
be presented. The cut-off bootstrap percentage is
50; that for the jackknife is 63%. The ITS matrix
consisted of 76 taxa, 74 American species and two
species of Pancratium as outgroup taxa. Pancratium
is part of the Eurasian clade that is the sister group
to the American Amaryllidaceae (Ito et al. 1999;
Meerow et al. 1999). Other members of the Eurasian
clade were also used as outgroup taxa, alone or to-
gether, without any topological changes in the
American clades, but the Pancratium sequences pre-
sented the least alignment ambiguities. Bremer
(1988) decay indices (DI) were calculated using the
program Tree Rot (Sorenson 1996). One hundred
replicate heuristic searches were implemented for
each constraint statement postulated by Tree Rot,
saving 10 trees per replicate.
The matrix was analyzed with gaps coded as
missing data but a gap matrix was also constructed
from the alignment using the program PAUPGAP
(Anthony Cox, RBG Kew), which applies a strict
interpretation of gaps (i.e., no partial homology).
This binary matrix was added to the alignment and
the combined data set analyzed cladistically as pre-
viously detailed.
Gaps Coded as Missing Data. Of 697 included
base positions in the analyses, 347 were parsimony
informative. Our initial 1000 replications with equal
weights applied found 9600 equally most parsi-
monious trees of length 51458 with a consistency
index (CI) of 0.54 and a retention index (RI) of 0.82.
SW produced at least 5000 equally parsimonious
trees with a length of 553270 (Fitch length 51459),
aCI50.73 (Fitch 50.54), and RI 50.91 (Fitch 5
0.82). The additional step of the SW trees is essen-
tially the ‘‘cost’’ of optimizing consistent characters
over highly homoplastic base positions (Lledo´etal.
1998; Meerow et al. 1999). Both the SW and Fitch
topologies were very similar; however, as the con-
sensus of the SW trees was slightly more resolved
than that of the Fitch trees, they will be the focus
of the discussion.
Two large subclades are resolved in the mono-
phyletic American Amaryllidaceae (Figs. 1, 2). The
first, or ‘‘hippeastroid’’ clade (Fig. 1) is referred to
as the diploid (n511, 12 or less primarily, though
polyploid species do occur), primarily extra-An-
dean element of the family (though several of the
genera have Andean representatives), comprising
the genera treated as the tribe Hippeastreae in most
recent classifications (Dahlgren et al. 1985; Mu¨ ller-
Doblies and Mu¨ ller-Doblies 1996; Meerow and Snij-
man 1998). The second subclade constitutes the tet-
raploid-derived (n523), Andean-centered tribes.
In both subclades, one tribe is sister to all of the
other genera in their respective groups. The Eus-
tephieae resolve as sister to the rest of the Andean
tribes (Fig. 2), but with weak support (DI 52, 56%
. 1. One of 5000 equally most parsimonious trees found by SW heuristic search of ITS sequence matrix for the
American Amaryllidaceae. Only the outgroup and ‘‘hippeastroid’’ clade is shown in detail; the tree is continued in Fig.
2. Numbers above branches are branch lengths; numbers below branches are DI (italic), bootstrap support (SW, under-
line), and jackknife support (equal weights, boldface). A white bar across a branch signifies a collapsed node in the
strict consensus of all trees. An exclamation point (!) in the support values indicates that both bootstrap and jackknife
. 2. One of 5000 equally most parsimonious trees found by SW heuristic search of ITS sequence matrix for the
American Amaryllidaceae. Only the ‘‘Andean’’ clade is shown in detail; the tree is continued in Fig. 1. Numbers above
branches are branch lengths; numbers below branches are DI (italic), bootstrap support (SW, underline), and jackknife
support (equal weights, boldface). A white bar across a branch signifies a collapsed node in the strict consensus of all
trees. An exclamation point (!) in the support values indicates that both bootstrap and jackknife 5100%.
bootstrap, no jackknife). In the hippeastroid clade,
the Griffineae (Worsleya and Griffinia) are sister to
the rest of the genera, with DI 51, 75% bootstrap,
and no jackknife support.
HIPPEASTROID CLADE. The Griffineae are a
well supported clade (DI 56, 94% bootstrap and
93% jackknife). Phycella is sister to the remaining
hippeastroids (DI 511, 99% bootstrap, 92% jack-
knife), followed by two species of Rhodophiala (DI 5
4, 93% bootstrap, 99% jackknife), with strong sup-
port only in the SW trees (100% bootstrap, DI 5
20). Two larger clades are then resolved, but with
only DI 51, 70% bootstrap and no jackknife sup-
port. The first unites two additional species of Rho-
dophiala and Hippeastrum blumenavium in a weakly
supported subclade (DI 50, 61% bootstrap and no
jackknife support) that is in turn sister to a well-
supported subclade of Zephyranthes species from
North America, South America and the West Indies
(DI 53, 97% bootstrap and 88% jackknife). The
overall support for this clade is strong (DI 53, 98%
bootstrap, 90% jackknife).
The largest clade of hippeastroid genera is split
into two smaller monophyletic groups in the SW
topology (Fig. 1). [In the consensus of our original
9600 trees with equal weights imposed, these two
subclades form a trichotomy with the heteroge-
neous clade described previously (not shown)]. The
smaller of the two represents the genus Hippeas-
trum, with the exception of H. blumenavium, and in-
cludes a species described as a Rhodophiala from
Brazil (all other species in this genus are from Chile
and Argentina), R. cipoana. Support for this clade
is strong (DI 53, 96% bootstrap, 93% jackknife).
Hippeastrum reticulatum is resolved as sister to all
other species in this clade. The second subclade es-
sentially represents what has been called tribe Ze-
phyrantheae (Traub 1963) or subtribe Zephyranthi-
nae (Mu¨ ller-Dobl ies and Mu¨ ller-Doblies 1996), ex-
cluding the Zephyranthes species that resolve else-
where. This clade, in which Sprekelia is resolved as
sister to the rest of the group, has only modest sup-
port (DI 51, 86% bootstrap, 76% jackknife). After
Sprekelia branches, a more strongly supported clade
(DI 53, 89% bootstrap, 87% jackknife) is resolved
consisting of two smaller subclades. The first is a
very well supported group of entirely Mexican Ze-
phyranthes (including Cooperia) species (DI 511,
100% bootstrap and jackknife). The second is a
weakly supported (DI 52, 64% bootstrap, 64%
jackknife) subclade that unites two additional South
American Zephyranthes species as sister group to
Habranthus, the latter within which is embedded the
genera Haylockia (5Zephyranthes pusila)andPyro-
lirion. The Habranthus subclade has no support and
ANDEAN CLADE. The Eustephieae are well
resolved (DI 57, 98% bootstrap, 95% jackknife,
Fig. 2) as a clade distinct from Stenomesseae,
though the sister group relationship of this clade
to the rest of the Andean clade is weak (DI 52,
56% bootstrap, no jackknife). Hymenocallideae is
also well-supported as a clade (DI 510, 99% boot-
strap, 100% jackknife), though only Hymenocallis is
resolved as monophyletic (DI 54, 100% bootstrap,
98% jackknife). The Hymenocallideae is sister to a
clade comprising some elements of Stenomesseae
(sensu Meerow and Snijman 1998) with DI 55,
strong bootstrap (93%) but weak jackknife (64%)
The petiolate-leafed Andean genera form a well-
supported clade (DI 58, 99% bootstrap, 89% jack-
knife) comprising elements of Stenomesseae and all
of Eucharideae (sensu Meerow and Snijman 1998).
However, only some of the component genera are
resolved. Phaedranassa is well supported (DI 510,
100% bootstrap and jackknife) and Eucrosia, less the
mesophyte E. dodsonii, is weakly supported (DI 5
2, 74% bootstrap, 70% jackknife). Stenomesson (pet-
iolate-leafed) also is resolved with weak support
(DI 51, 69% bootstrap and jackknife). The two spe-
cies of Rauhia form a weak clade in only the boot-
strap consensus (54%). By contrast, relationships
among the genera of Eucharideae sensu Meerow
and Snijman (1998) are poorly resolved.
Gap Matrix Added to Alignment. Of 797 total
characters, 396 were parsimony informative. Our
initial 1000 replicate heuristic search with equal
weights applied found at least 7950 trees of length
51644, with CI 50.54 and RI 50.82. SW found
198 trees of length 5634628 (Fitch length 51644),
with CI 50.75 (Fitch 50.54) and RI 50.91 (Fitch
As with the analysis of base substitutions alone
(Figs. 1, 2), the hippeastroid and Andean subclades
are resolved, with Griffineae and Eustephieae as
sister to each, respectively (Figs. 3, 4). Bootstrap
support for both subclades is higher than with the
base substitution matrix alone (84% for the hip-
peastroid group, 90% for the Andean group), but
only the Andean subclade receives jackknife sup-
port (92%). DI 55 for the Andean group, but only
1 for the hippeastroid subclade.
There are several consistent and significant gaps
in our alignment that represent synapomorphies for
the Andean clade (Fig. 5), and these strengthened
the support values for this group: 1) base position
(bp) 43–52, bp 90–99 in ITS1 and 22–28 in ITS2 (ex-
cept Eustephieae).
HIPPEASTROID CLADE. When the gap matrix
is included in the analysis (Fig. 3), resolution of
these taxa is very similar to that of the base sub-
stitution matrix alone (Fig. 1). Polyphyletic Zephyr-
anthes and Rhodophiala are once again indicated, and
Phycella is sister to the rest of the Hippeastreae.
ANDEAN CLADE. The SW trees from the
. 3. One of 196 equally most parsimonious trees found by SW heuristic search of ITS sequences plus gap matrix
for the American Amaryllidaceae. Only the ‘‘Andean’’ clade is shown in detail; the tree is continued in Fig. 4. Numbers
above branches are branch lengths; numbers below branches are DI (italic), bootstrap support (SW, underline), and
jackknife support (equal weights, boldface). A white bar across a branch signifies a collapsed node in the strict consensus
of all trees. Black bars and lower case letters refer to synapomorphic indels illustrated in Fig. 5. An exclamation point
(!) in the support values indicates that both bootstrap and jackknife 5100%.
combined base substitution and gap matrix did add
further resolution to the Andean clade, though the
overall topology is not changed. Within the Hy-
menocallideae, Ismene and Leptochiton are resolved
as sister to Hymenocallis, though Ismene is still not
monophyletic unless Leptochiton is included. Sup-
port is weak for this group (DI 50, 54% bootstrap).
In the petiolate-leafed clade, Rauhia is resolved as a
. 4. One of 196 equally most parsimonious trees found by SW heuristic search of ITS sequence matrix alone for
the American Amaryllidaceae. Only the ‘‘Andean’’ clade is shown in detail; the tree is continued in Fig. 3. Numbers
above branches are branch lengths; numbers below branches are DI (italic), bootstrap support (SW, underline), and
jackknife support (equal weights, boldface). A white bar across a branch signifies a collapsed node in the strict consensus
of all trees. Black bars and lower case letters refer to synapomorphic indels illustrated in Fig. 5. An exclamation point
(!) in the support values indicates that both bootstrap and jackknife 5100%.
monophyletic sister group to Phaedranassa but with
weak support (bootstrap 560%). Rauhia itself is
monophyletic, however (DI 50, 94% bootstrap,
73% jackknife). The Eucharideae sensu Meerow and
Snijman (1998) is again dispersed throughout the
petiolate-leafed clade.
Hippeastroid Clade and the Position of
Griffineae. All but two of the genera treated by
Meerow and Snijman (1998) as part of Hippeas-
treae are resolved as a well supported monophy-
. 5. Significant insertion/deletions in the ITS region
that mark the ‘‘hippeastroid’’ and Andean clades, respec-
tively, of the American Amaryllidaceae.
letic group in all the analyses (Figs. 1, 3). The two
genera that lie outside of this clade are Worsleya and
Griffinia, both Brazilian endemics, exhibiting the
rare character of blue pigmentation in the flowers.
In our previous combined plastid analysis (Meerow
et al. 1999), both are positioned within the Ameri-
can clade, but unresolved with either Hippeastreae
s.s. or a weakly supported tetraploid Andean clade.
ITS strongly resolves Worsleya and Griffinia as sister
genera, both resolved on long branches. Ravenna
(1974) hinted at a perceived relationship between
these genera, though he did not include Worsleya in
his concept of Griffineae, and further allied this
tribe with Amaryllideae (as Crineae). Clearly, Mu¨ ll-
er-Doblies and Mu¨ ller-Doblies’s (1996) placement of
Worsleya in Hippeastrum is not supported by this
study. Within the hippeastroid clade, recognition of
a distinct tribe Griffineae appears well justified
(Figs. 1, 3) with strong bootstrap and jackknife sup-
port (95%). Together these Brazilian endemics
would appear to represent an isolated Brazilian
shield element of the family’s diversification in the
neotropics. Bootstrap support for the sister group
status of Griffineae to Hippeastreae is only mod-
erate, however (75 or 84%, Figs. 1, 3), and the lack
of jackknife support indicates that this resolution is
only supported when the matix is successively
In a survey of internal morphology of American
and African Amaryllidaceae, Arroyo and Cutler
(1984) noted that all American species surveyed
have scapes with obvolute bracts. All African tribes
have equitant bracts (in Agapanthus, sister to Amar-
yllidaceae, the bracts are fused along one side).
Meerow et al. (1999) reported that Calostemmateae
(Calostemma and Proiphys), Lycoris, and Pancratium
species with free bracts show the equitant condi-
tion. We only recently confirmed that Worsleya has
a solid scape (like Griffinia), but equitant bracts like
the rest of the family (in Griffina the bracts are fused
on one side). In both analyses, the sister relation-
ship of Griffineae and Hippeastreae has a DI of
only 1; i.e., in trees only one step longer than those
shown in Figs. 1 and 3, the position of Griffineae
is unresolved within the American group. Thus
there is a lingering possibility that Griffineae is the
most ancient lineage in the Americas, and possibly
sister to both Hippeastreae and the Andean clade.
The Chilean genus Phycella is next resolved as sis-
ter to the rest of tribe Hippeastreae (Fig. 3). The
next clade resolved is also composed of species
with a southern displacement from the equator, two
species of Rhodophiala, one Chilean and one from
Argentina. Rhodophiala is the first of several Amer-
ican genera that ITS indicates are polyphyletic. An-
other two Chilean species of Rhodophiala resolve in
a grade as part of a weakly supported clade in-
cluding Hippeastrum blumenavium. Rhodophiala ci-
poana, from Minas Gerais, Brazil, a considerable
disjunct from the rest of genus, is firmly nested
with Hippeastrum, strongly suggesting a misdiag-
nosis of this small-flowered, dwarf species’ affilia-
tion by Ravenna (1970). In fact, the species has 2n
522 chromosomes (J. Dutilh, pers. comm.), a num-
ber unknown in Rhodophiala, (Flory 1977) but char-
acteristic of Hippeastrum (Naranjo and Andrada
1975). Rhodophiala has at times been treated as part
of Hippeastrum (e.g., Baker 1888; Bentham and
Hooker f. 1883; Traub and Moldenke 1949), but few
specialists in the family question Traub’s later
(Traub 1952, 1953) segregation of the genus. The
limits of the genus have continued to be controver-
sial, however (e.g., Hunziker 1985). Two chromo-
some groups have been reported in the genus, 2n
516 and 2n 518 (Ficker 1951; Flory 1977; Naranjo
1969; Satoˆ 1942). Rhodophiala bifida has 2n516 chro-
mosomes (Naranjo 1969); no published reports ex-
ist for R. bagnoldii. While our data leaves little doubt
that Rhodophiala is distinct from Hippeastrum, its ap-
parent polyphyly is puzzling and requires wider
sampling of the genus (but see discussion below).
It should be noted that the resolution of the Rho-
dophiala bifida/R. bagnoldii clade as sister to the re-
maining Hippeastreae has no jackknife support, in-
dicating that this resolution is only well-supported
when highly homoplasious base positions are
Hippeastrum is inarguably monophyletic if Rho-
dophiala cipoana is included (Figs. 1, 3) with the ex-
ception of a single species, H. blumenavium. Hip-
peastrum blumenavium is an unusual species mor-
phologically (petiolate leaves; few, wedge-shaped
seeds with an elaisome). It was first described as
Griffinia blumenavia Koch and Bouche ex Carr. ITS
sequences indicate that it is much more closely re-
lated to Rhodophiala than to Hippeastrum. Satoˆ (1938)
reported a chromosome number of 2n577 for this
species, while Arroyo (1981) found 2n520. Inde-
pendent unpublished studies (G. Smith pers.
comm.; J. Dutilh pers. comm) report 2n518, the
most common chromosome number in Rhodophiala
Ficker 1952; Flory 1977; Naranjo 1969; Satoˆ 1942).
The sister relationship to two Rhodophiala species is
only weakly supported (bootstrap 564%, DI 50),
and ultimately this species may be best treated as
a monotypic genus.
Hippeastrum reticulatum resolves as sister to the
rest of the genus. Endemic to the understory of At-
lantic rain forest of Brazil, H. reticulatum exhibits
novel fruit and seed morphology. The seeds are
round and turgid, and the dehisced capsule is
bright red on the inner surface, perhaps functioning
mimetically. As the next two terminal species of
Hippeastrum are also Brazilian, a Brazilian origin for
the genus seems likely. Further species relation-
ships in this genus require additional sequence
data; we originally sampled nearly 3 dozen species,
but ITS base substitution rates are too low for ad-
equate resolution beyond what is presented in Figs.
2and4.Hippeastrum parodii is endemic to Argen-
tina, and H. molleviquensis (Bolivia) and H. macbridei
(Peru) represent the Andean species. Though Hip-
peastrum is monophyletic, excluding H. blumena-
vium, the sister relationship of the genus to one
clade of Zephyranthinae lacks any support in our
trees (Fig. 2), and is unresolved by the sequence
only matrix (Fig. 1).
Tribe Zephyrantheae (Traub 1963; Traub also in-
cluded Rhodophiala in this tribe) or Hippeastreae
subtribe Zephyranthinae (Mu¨ller-Doblies and
Mu¨ ller-Doblies 1996) are clearly polyphyletic. At
present this is largely due to the apparent poly-
phyly of Zephyranthes itself, a genus which Meerow
(1995) suggested may represent convergence of two
distinct (albeit related) lineages in North (Meso)
America and South America. Our wider sampling
of this genus now indicates a putative triple origin
for this genus. A most unexpected result is the po-
sition of the two species of this genus (Z. atamasco
and Z. simpsonii) from the southeastern United
States. Rather than appearing as sister to the well
supported Mexican clade, Z. atamasco and Z. simp-
sonii are nested within the clade of South Ameri-
can/West Indian species (very well supported) that
is sister to the Rhodophiala/Hippeastrum blumenavium
clade. Sprekelia, a Mexican endemic (anecdotal re-
ports of an Andean species have never been docu-
mented) with a highly zygomorphic perianth
adapted for hummingbird pollination (two Hip-
peastrum species, H. cybister and H. angustifolium,
have evolved similar perianth) resolves with mod-
erate support as sister to the rest of Zephyranthi-
neae (Fig. 1, 3). The latter clade consists of a para-
phyletic Habranthus (with no support and a DI 51)
as a weakly supported sister to two South Ameri-
can Zephyranthes (Z. flavissima and Z. mesochloa), and
a well-supported Mexican Zephyranthes clade. This
heterogeneous subclade of Zephyranthinae has
weak internal support, however. The segregate ge-
nus Cooperia (crepuscular, long-tubed Zephyran-
thes; represented in this study by Z. smallii and Z.
drummondii) is nested within the Mexican Zephyr-
anthes clade, the internal resolution of which is poor.
Both Pyrolirion and Haylockia (5Zephyranthes pusil-
la) are embedded in Habranthus. These two genera
have actinomorphic flowers, while those of Habran-
thus are zygomorphic. If zygomorphy is indeed ple-
siomorphic in the family (Meerow and Snijman
1998), our data suggest that the change to actino-
morphy is relatively facile, and probably under sim-
ple genetic control.
The Zephyranthineae have been broadly catego-
rized as having a basic chomosome number of x 5
6 (Flory 1968, 1977), but considerable variation is
exhibited among the species of Zephyranthes and
Habranthus, the two largest genera, including eu-
ploid and aneuploid series. There is little consisten-
cy within the species of each of the three Zephyr-
anthes clades resolved by ITS; at least some of the
species within each has 2n 524 chromosomes. If
the results of this analysis are accurate, then chro-
mosome number as well as morphology has
evolved in parallel among the three subclades. An
alternative hypothesis might be that reticulation
early in the evolution of Zephyranthes and its allies
completely obscures the phylogenetic relationships
(Doyle 1992; McDade 1990, 1992, 1995). Intergener-
ic hybrids among genera of the Zephyranthinae are
not unknown (Traub 1952, 1963; Flory 1968). The
considerable amount of homoplasious base substi-
tutions within the Hippeastreae (evidenced by the
lack of internal branch support in trees resolved by
the unweighted matrices) may be an indication that
reticulation did occur in this clade. We plan to in-
vestigate this further by looking for corroboration
of this pattern with plastid sequence data, as well
as with tests for recombination within sequence
Andean Clade. Eustephieae are well resolved
with very strong support as distinct from Steno-
messeae (Figs. 2, 4) as argued by Meerow (1995).
Mu¨ ller-D oblies and Mu¨ller-Doblies (1996) followed
Dahlgren et al. (1985) in placing this tribe within
Stenomesseae. Support for this tribe as sister to the
rest of the Andean clade rises from weak in the
sequence topology (Fig. 2) to very strong in the se-
quences plus gaps trees (Fig. 4). This is a striking
parallel to the situation in the hippeastroid clade,
where a small clade is sister to all of the rest of the
genera in their respective groups, though the po-
sitioning of Eustephieae is much stronger than the
analogous relationship of Griffineae to the Hip-
peastreae (Figs. 1, 3). The Eustephieae represents
the southern limits of the tetraploid Andean clade,
occurring in the southern Andes of Peru, and the
northern Andes of Argentina, Bolivia and Chile.
They possess the putatively plesiomorphic condi-
tion of leaf mesophyll palisade (Meerow and Snij-
man 1998), unlike any other genus in the Andean
clade. Support for the internal resolution of Euste-
phieae is also strong.
The fleshy seeded Hymenocallideae are also
well-supported as a tribe distinct from Eucharideae
(bootstrap 599%). In fact, the affinities of Hymen-
ocallideae are more with the lorate-leafed Steno-
messeae (dry, flat, obliquely winged seed) than Eu-
charideae (turgid, oily seed). The infratribal rela-
tionships of Hymenocallideae are ambiguous, how-
ever, except for the strong monophyly of
Hymenocallis. The ambiguity surrounds the relation-
ships of the Andean genera Ismene and Leptochiton.
The combined bp and gaps matrix very weakly re-
solves monophyly for these Andean taxa, but ren-
ders Ismene polyphyletic. Hymenocallis has only 2–3
species in South America, and represents the North
and Meso-American vicariant of the tribe. Leptochi-
ton and the three subgenera of Ismene (Ismene, Eli-
sena, Pseudostenomesson) are endemic to the central
Andes. Leptochiton is the only member of the tribe
with phytomelanous seeds and also has the lowest
chromosome number [2n524 according to Snoad
(1952); 2n534 in our accessions (Meerow, unpub-
lished data)]. Plastid sequences resolve Ismene as
monophyletic (Meerow et al. 1999), but with weak
ITS was most helpful in elucidating the relation-
ships of the tribes Stenomesseae and Eucharideae,
two tribes which showed insufficient plastid DNA
divergence to resolve their phylogenetic relation-
ships (Meerow et al. 1999). Most surprising is the
resolution of a petiolate-leafed clade containing el-
ements of both Eucharideae and Stenomesseae
[(100% bootstrap, 93% jackknife, DI 57 with both
bp and gaps (Fig. 4)]. Despite the fact that petiolate
leaves have evolved independently several times
elsewhere in Amaryllidaceae (Amaryllideae, Calos-
temmateae, Griffineae, Haemantheae, Hymenocal-
lideae, Hippeastreae), the molecular data indicate
that it is a synapomorphy for this clade. Whereas
Meerow (1987) pointed out the probable paraphyly
of Stenomesson, suggesting that the petiolate-leafed
species of that genus might be ancestral to Rauhia,
Phaedranassa, and Eucrosia (and the extinct Mathieua
galanthoides), ITS sequences clearly indicate that
Stenomesson is polyphyletic, and that petiolate
leaves evolved only once in the Andean Amarylli-
daceae. All of the petiolate-leafed Stenomesseae are
more closely related to Eucharideae sensu Meerow
and Snijman (1998) than to the lorate-leafed Sten-
omesseae. The sampling of this clade is still incom-
plete, however, with only Stenomesson s.s., Phaedran-
assa, and Rauhia resolving as monophyletic (Rauhia
only when the gap matrix is included) with good
bootstrap support for the latter two when the gaps
matrix is included in the analysis (Fig. 4). The sister
relationship of Rauhia and Phaedranassa resolves
only in the combined bp and gap matrix. While
three species of Eucrosia form a well-supported
clade, the relationships of the only mesophyte in the
genus, E. dodsonii, are unresolved. Although seed
morphology for this species is unknown, ovule
morphology is more like that of Caliphruria, Euchar-
is, and Urceolina (Eucharideae), all of which have
globose ovules, than the winged, flattened ovules
of Stenomesseae sensu Meerow and Snijman (1998).
A further link of E. dodsonii to these genera is the
rain forest understory habitat that all of them share.
Meerow and Dehgan’s (1985) placement of this spe-
cies in Eucrosia may therefore require re-analysis
once we obtain a broader sampling of the petiolate
clade. Not only does Eucharideae sensu Meerow
and Snijman (1998) fail to resolve as a monophyletic
group, neither Eucharis nor Caliphruria appear
monophyletic. The lack of resolution in this group
may in part be attributable to a relatively recent
evolutionary history tied to the uplift of the Andes
in the Pliocene (van der Hammen 1974, 1979; Taylor
The lorate-leafed remnants of the Stenomesseae
form a weakly supported clade with the epiphytic
Pamianthe at the base. In trees 2 steps longer, Pa-
mianthe is unresolved relative to Hymenocallideae
or the lorate-leafed Stenomesseae. Paramongaia we-
berbaueri and Stenomesson mirabile (the latter on an
extremely long branch) are well supported sister
taxa (Fig. 2) in the SW (bootstrap 592%), less so
in the unweighted trees (jackknife 573%) from the
sequence matrix alone. With the gap matrix added
(Fig. 4), S. mirabile is sister to the rest of Stenomesson
but with only 70% bootstrap support in the SW
trees. As in Paramongaia, the free portion of the sta-
minal filament in S. mirabile is inserted below the
rim of the staminal cup.
Biogeographic and Evolutionary Implications.
Raven and Axelrod (1972), Arroyo and Cutler
(1984), and Meerow (1995) have interpreted the en-
demic American Amaryllidaceae as a classical
Gondwanaland disjunct from an African ancestor.
However, combined plastid sequence phylogenies
of the family (Meerow et al. 1999) indicate that the
sister group relationships of the American genera
are with the Eurasian tribes of the family, rather
than any African clade. Meerow et al. (1999) were
reluctant to suggest when or by what migration
path the family entered the Americas, but invoked
the Madrean-Tethyan hypothesis (Axelrod 1973,
1975) as a possibility. To note, a single species of
Pancratium, P. canariensis, part of the Eurasian sister
clade to the American Amaryllidaceae, is endemic
to the Canary Islands and could conceivably rep-
resent an isolated remnant of migrations across the
islands of the mid-Atlantic ridge in the late Creta-
ceous or early Paleocene times. The complete ab-
sence of a fossil record for Amaryllidaceae [except
for some recent pollen deposits from Brazil (Behl-
ing 1995)] makes it very difficult to assess any of
several hypotheses of the family’s history in America.
Within each of the two major clades of American
amaryllids, we have a parallel topology in the lower
branches. In both cases, a small clade resolves as sister
to a larger, more diverse clade. In the hippeastroid
clade, the ‘‘little’’ sister is the eastern Brazil endemic
Griffineae; in the Andean clade, it is the Eustephieae,
which occurs in its present distribution considerably
disjunct from the central Andean locus of diversity
for the tetraploid Andean amaryllids.
This pattern, and the low base substitution rates
apparent in the more terminal subclades of both
larger groups, might indicate that that the family
was subjected to a bottleneck in its evolution some
time after its initial colonization of the Americas.
Thus, one hypothesis of the history of the American
Amaryllidaceae could be that the family was re-
duced to two peripheral isolates from its former
range, a southern Andean group (Eustephieae) and
a Brazilian Shield element (Griffineae), from which
all modern diversity of the family in America later
evolved. Glaciation would be the most reasonable
factor to have been involved in reducing the former
range of the family in America. Both groups exhibit
a morphological plesiomorphy: leaf mesophyll pal-
isade in the Eustephieae, and equitant bracts in
Griffineae (Worsleya). They also exhibit a consider-
able amount of unique sequence in the ITS regions.
Alternatively, the Eustephieae and Griffineae could
merely represent early isolates from the rest of their
respective clades, with no particular claim of great-
er antiquity.
Within the larger hippeastroid subclade, the next
terminal clades are primarily displaced towards the
southern region of South America, with Phycella
(Chile) followed by two Rhodophiala species (Chile
and Argentina). The next clade to resolve is the
most geographically heterogeneous group in the
entire topology, including species from Brazil (Hip-
peastrum blumenavium, Zephyranthes cearensis), Ar-
gentina (Z. candida, Z. filifolia), the West Indies (Z.
rosea from Cuba), and North America (Z. atamasco
and Z. simpsonii). The broadly dispersed Zephyran-
thes subclade is particularly intriguing. If the ITS
phylogeny is accurate, the sister relationship of Z.
rosea to the North American species would suggest
that the genus entered the southeastern U.S. via
Cuba after migrating from South America.
The boreotropical hypothesis postulates that,
during the early Tertiary, biotic interchange oc-
curred between paleotropical floras of Eurasia and
North America over a direct land bridge or narrow
water gaps (Wolfe 1975; Tiffany 1985a, 1985b; Tay-
lor 1988; Lavin and Luckow 1993; Lavin 1995). Dur-
ing this time, little or no migration was possible
between North and South America (Raven and Ax-
elrod 1974). Fossil evidence has accumulated that
supports the existence of a diverse tropical flora in
North America up until the late Eocene or early
Oligocene when the North Atlantic widened and
temperate climates began to prevail in North Amer-
ica (Taylor 1988; Lavin 1995). That the American
and Eurasian Amaryllidaceae are sister groups
(Meerow et al. 1999) provides one congruent piece
of evidence to suggest that this model might fit the
Amaryllidaceae. However, a further expectation
from phylogenetic analysis would be that the South
American Amaryllidaceae should resolve as de-
rived from North American progenitors, as Lavin
(1995) has demonstrated for the legume tribe Ro-
binieae. This is not explicit in our trees (Figs. 1–4).
However, within the hippeastroid clade (Figs. 1, 3),
we have a well-supported clade of Zephyranthes spe-
cies that includes species endemic to North Amer-
ica, the West Indies, and South America, whereas
the other clades of Zephyranthes species show geo-
graphic congruence. If this resolution represents the
true phylogeny (versus the consequence of recom-
bination early in the evolution of the hippeastroid
clade), then North American species of Zephyranthes
may represent the remnants of an early North
American lineage in the family. It should be noted,
however, that the two North American species are
the terminal-most taxa in their clade, their sister
species is a West Indian taxon (Z. rosea), and all
three appear to be derived from South American
progenitors. As no Tertiary fossil record exists for
the Amaryllidaceae on any continent, we conclude
at this point that the evidence for a Boreotropical
origin for the American Amaryllidaceae is at best
ambiguous. Our data do not suggest that extant
South American Amaryllidaceae were derived from
North American progenitors, despite the sister
group relationship of the entire American clade to
the Eurasian (Meerow et al. 1999), since South
American taxa are in all of the basal positions in
our trees. Based on our sequence topologies, South
America seems the most likely center of origin for
the family’s extant diversity in the Americas, with
later migration to North America, probably across
the Panamanian Isthmus (Raven and Axelrod 1974;
Gentry 1982; Taylor 1990; Graham 1995), or, in the
case of Zephyranthes in the southeastern U.S., via is-
land hopping through the Caribbean. Given the
weak to moderate support for the Griffineae as sis-
ter to the Hippeastreae (versus an unresolved po-
sition in the American clade) only in our SW to-
pologies, and the symplesiomorphic scape bract
morphology of Worsleya, a reasonable hypothesis for
a migration pattern would have the progenitor en-
tering northern South America from North or trop-
ical Africa. A land bridge did exist between the
Brazilian Bulge and Nigeria until the end of the
Cretaceous (Rand and Mabesoone 1982). This same
ancestor may have been shared with the Eurasian
sister clade of the American group. The American
species of Crinum, the only pantropical amaryllid
genus, are nested within a clade of North and trop-
ical African species of the genus by ITS sequences
(unpub. data), suggesting a similar pattern for the
independent entry of this group into America. In
any event, direct evidence for this or any other hy-
pothesized pathway for Amaryllidaceae is lacking.
In conclusion, ITS sequences provide a well-re-
solved phylogeny of the monophyletic American
Amaryllidaceae, but the early origins of the family
in America remain ambiguous. The genera resolve
into two subclades, an Andean-centered, tetra-
ploid-derived group, and a clade conforming to the
Hippeastreae sensu Meerow and Snijman (1998). In
both cases, a small, geographically constrained
South American tribe is sister to the more diverse
and dispersed elements of the group. At present,
the resolution of the Griffineae as part of the hip-
peastroid subclade receives moderate support only
with SW weighting of the data matrices. The reso-
lution of Eustephieae as sister to rest of the Andean
genera is strongly supported when sequence gaps
are included as part of the data matrix analyzed.
The Griffineae and Eustephieae represent either an-
cient or isolated lineages within the American clade
of the family. Within the hippeastroid clade, the ap-
parent polyphyly of two genera, Rhodophiala and
Zephyranthes, for which any other support is lack-
ing, opens the possibility that reticulation might
have occurred in the early evolution of the lineage.
This issue requires further investigation.
Taxonomic Implications. At present, it is pre-
mature to suggest any taxonomic changes within
the hippeastroid clade of our ITS phylogeny, other
than recognition of the tribe Griffineae as distinct.
However, within the Andean clade ITS indicates
that the tribe Stenomesseae is polyphyletic, a reso-
lution that was also suggested by plastid sequences
(Meerow et al., 1999). A distinct petiolate-leafed
clade, comprising the Eucharideae and the petiolate
genera of the Stenomesseae, is strongly supported.
As the type species of Stenomesson (the oldest name
in the clade) is a petiolate-leafed species [S. flavum
(R. & P.) Herb.], this clade must bear the name Sten-
omesseae. ITS also strongly supports recognition of
Hymenocallideae as a distinct tribe that is sister to
the lorate-leafed remnants of the Stenomesseae. The
earliest prior name for the lorate-leafed species of
Stenomesson is Clinanthus Herbert (1821) and the
name Clinantheae is proposed for the new tribe. It
should be noted that the ITS sequence topology
without gaps resolves Stenomesson mirabile (and S.
viridiflorum by inference) as sister to Paramongaia
rather than the rest of the lorate-leafed Stenomesson
species (Fig. 2). The latter resolution occurs when the
gap matrix is included (Fig. 4), but is not as well
supported as the former. We feel it is best to treat
these two erstwhile Stenomesson species as part of
Clinanthus and retain Paramongaia as a monotypic ge-
nus until the clade is more fully sampled.
Clinantheae Meerow, tribus novus.
Tribus novus herbarum perennium bulbosarum
andinarum, foliis loratis vel linearibus, floribus cu-
pula staminea, et seminibus phytomelanis com-
planatis alatis.
Bulbous perennial herbs, terrestrial, rarely epi-
phytic. Bulbs tunicate, often forming a long neck.
Leaves linear or lorate, hysteranthous or synan-
thous, rarely persistent, often glaucous, lacking pal-
isade in the mesophyll. Inflorescence scapose, pseu-
doumbellate (reduced helicoid cymes), the scape
solid, terminated by two obvolute spathe bracts that
enclose the flowers in bud, bracts rarely fused into
a tube. Perianth crateriform, funnelform tubular, tu-
bular, or campanulate, often brightly colored, ped-
icellate, each subtended by a bracteole, consisting of
six tepals in two series fused below into a tube of
varying length. Stamens 6, the filaments fused be-
low into a staminal cup, free filament inserted at or
below the rim of the cup. Stigma capitate; ovary
inferior, trilocular, ovules numerous, axile in pla-
centation, compressed. Fruit a papery or woody
loculicidal capsule; seeds dry, flattened, obliquely
winged with a black or brown phytomelanous tes-
ta. 2n546.
TYPE: Clinanthus luteus Herbert. App. Edwards’s
Bot. Reg. 40 (1821).
Genera: Clinanthus (ca. 20), Pamianthe (2) Para-
mongaia (1), Pucara (1).
The following transfers of currently recognized
linear or lorate-leafed Stenomesson species to Clinan-
thus are necessary. Only the basionym is provided,
pending full synonomy in a monograph of the ge-
Clinanthus callacallensis (Ravenna) Meerow, comb.
Stenomesson callacallense Ravenna. Pl. Life 30: 76
Clinanthus campodensis (Ravenna) Meerow, comb.
Stenomesson campodense Ravenna. Pl. Life 27: 75
Clinanthus caracensis (Ravenna) Meerow, comb.
Stenomesson caracense Ravenna. Pl. Life 30: 76 (1974).
Clinanthus chihuanhuayu (Ca´rdenas) Meerow,
comb. nov.
Haylockia chihuanhuayu Ca´rdenas, Pl. Life 29: 44
Clinanthus coccineus (R. & P.) Meerow, comb. nov.
Pancratium coccineum R. & P., Fl. Peruv. 3: 54 (1802).
Clinanthus croceus (Savigny) Meerow, comb. nov.
Pancratium croceum Savigny, Lam. Encycl. 4: 725
Clinanthus elwesii (Baker) Meerow, comb. nov.
Callithauma virdiflorum (R. & P.) Herb. var. elwesii
Baker, Gard. Chron., n.s. 9: 756 (1888).
Clinanthus flammidis (Ravenna) Meerow, comb.
Stenomesson flammidum Ravenna, Pl. Life 27: 73
Clinanthus fulvus (Herbert) Meerow, comb. nov.
Coburghia fulva Herbert, Edward’s Bot. Reg. 18:
t.1497 (1832).
Clinanthus glareosus (Ravenna) Meerow, comb.
Stenomesson glareosum Ravenna, Pl. Life 27: 73
Clinanthus humilis (Herbert) Meerow, comb. nov.
Clitanthes humilis Herbert, Edwards’s Bot. Reg. 25:
Misc. 87 (1839).
Clinanthus imasumac (Vargas) Meerow, comb. nov.
Stenomesson imasumac Vargas, Biota 8: 38 (1969).
Clinanthus incarnatus (H. B. K.) Meerow, comb.
Pancratium incarnatum H.B.K., Nov. Gen. & Sp. 1:
280 (1816).
Clinanthus incarus (Kraenzlin) Meerow, comb. nov.
Stenomesson incarum Kraenzlin, Bot. Jahrb. 40: 238
Clinanthus macleanicanus (Herbert) Meerow,
comb. nov.
Clitanthes macleanica Herbert, Edwards’s Bot. Reg.
25: misc. 87 (1839).
Clinanthus microstephus (Ravenna) Meerow, comb.
Stenomesson microstephium Ravenna, Pl. Life 34: 76
Clinanthus mirabilis (Ravenna) Meerow, comb.
Stenomesson mirabile Ravenna, Pl. Life 27: 77 (1971).
Clinanthus recurvatus (R. & P.) Meerow, comb. nov.
Pancratium recurvatum R. & P., Fl. Peruv. 3: 54 (1802).
Clinanthus sunchubambae (Ravenna) Meerow,
comb. nov.
Stenomesson sunchubambae Ravenna, Onira 1(2): 17
Clinanthus variegatus (R. & P.) Meerow, comb. nov.
Pancratium variegatum R. & P., Fl. Peruv. 3: 55 (1802).
Clinanthus viridiflorus (R. & P.) Meerow, comb.
Pancratium viridiflorum R. & P., Fl. Peruv. 3: 55
. Some of the sequences were
generated at the DNA Sequencing Core of the Interdisci-
plinary Center for Biotechnology Research at the Univer-
sity of Florida. Financial support was provided in part by
NSF grants DEB-968787 (AWM and CLG) and IBN-
9317450 (CLG), and a University of Florida Institute of
Food and Agricultural Sciences Research Enhancement
Grant. We thank the many individuals and institutions
that provided access to their living collections or collected
leaf samples. Gerald L. Smith and an anonymous reviewer
provided useful reviews of this paper.
. 1984. Evolutionary and
taxonomic aspects of the internal morphology in
Amaryllidaceae from South America and Southern
Africa. Kew Bulletin 39: 467–498.
, D. I. 1973. History of the Mediterranean ecosys-
tem in California. Pp. 225–227 in Mediterranean type
ecosystems: origin and structure, eds. F. di Castri and
H. A. Mooney Berlin: Springer-Verlag.
———. 1975. Evolution and biogeography of Madrean-
Tethyan scleophyll vegetation. Annals of the Missouri
Botanical Garden 62: 280–334.
, J. G. 1888. Handbook of the Amaryllideae. London:
George Bell and Sons.
, B. G. 1991. Nuclear ribosomal DNA phyloge-
netics of the tarweeds (Madiinae: Asteraceae). Amer-
ican Journal Botany 78(suppl., abstract): 165–166.
———. 1992. Phylogenetic utility of the internal tran-
scribed spacers of nuclear ribosomal DNA in plants:
an example from the Compositae. Moleccular Phylo-
genetics and Evolution 1: 3–16.
———, M. J. S
, and M. J. D
1995. The ITS region of nuclear ribosomal DNA: a
valuable source of evidence on angiosperm phyloge-
ny. Annals of the Missouri Botanical Garden 82: 247–
, R. M., A. M. P
, and M. W. C
. 1997.
Phylogenetics of subtribe Orchidinae (Orchidoideae,
Orchidaceae) based on nuclear ITS sequences. 2. In-
frageneric relationships and reclassification to achieve
monophyly of Orchis sensu stricto. Lindleyana 12:
, H. 1995. Investigations into the late Pleistocene
and Holocene history of vegetation and climate in
Santa Catarina (S Brazil). Vegetation History and Ar-
chaeobotany 4: 127–152.
. 1883. Genera Plantarum,
volume 3. London: L. Reeve and Co.
, K. 1988. The limits of amino acid sequnce data in
angiosperm phylogenetic reconstruction. Evolution
42: 198–213.
F. W
. 1995. Phylogenetic relationships
in maloideae (Rosaceae): evidence from sequences of
the internal transcribed spacers of nuclear ribosomal
DNA and its congruence with morphology. American
Journal Botany 82: 903–918.
, a n d P. F. Y
. 1985.
The families of monocotyledons: structure, evolution, and
taxonomy. Berlin: Springer-Verlag.
, J. P., A. M. P
P. L
, and M. W. C
. 1999. Molecular phylo-
genetics of Diseae (Orchidaceae): a contribution from
nuclear ribosomal ITS sequences. American Journal
Botany 86: 887–899.
, J. J. 1992. Gene trees and species trees: molecular
systematics as one-character taxonomy. Systematic
Botany 17: 144–163.
——— and J. L. D
. 1987. A rapid DNA isolation pro-
cedure from small quantities of fresh leaf tissue. Phy-
tochemical Bulletin 19: 11–15.
, T. and M. J. D
. 1997. Phylogenetic re-
lationships of Sambucus and Adoxa (Adoxoideae,
Adoxaceae) based on nuclear ribosomal ITS sequenc-
es and preliminary morphological data. Systematic
Botany 22: 555–573.
, J. S. 1969. A successive approximations approach
to character weighting. Systematc Zoology 18: 374–
———, V. A. A
G. K
. 1996. Parsimony jackknifing outperforms
neighbor-joining. Cladistics 12: 99–124.
. 1996. Resurrection of Them-
idaceae for the Brodiaea alliance, and recircumscrip-
tion of Alliaceae, Amaryllidaceae and Agapanthoi-
deae. Taxon 45: 441–451.
, T. 1951. Chromosomes of two narrow-leaved Am-
aryllis species, and the generic type species of Ama-
ryllis belladonna L. Plant Life 7: 68.
, W. M. 1971. Toward defining the course of evolu-
tion: Minimum change for a specific tree topology.
Systematic Zoology 20: 406–416.
, W. S. 1968. Chromosome diversity in species, and
in hybrids, of tribe Zephyrantheae. Nucleus 11: 79–
———. 1977. Overview of chromosomal evolution in the
Amaryllidaceae. Nucleus 20: 70–88.
, A. 1982. Neotropical floristic diversity: phytogeo-
graphical connections between Central and South
America, Pleistocene climatic fluctuations, or an ac-
cident of the Andean orogeny? Annals of the Missou-
ri Botanical Garden 69: 557–593.
, A. 1995. Development of affinities between Mex-
ican/Central American and northern South america
lowland and lower montane vegetation during the
Tertiary. Pp. 11–22 in Biodiversity and conservation of
neotropical montane forests eds. S. P. Churchill, H. Bal-
slev, E. Forero, and J. L. Luteyn. New York: New York
Botanical Garden.
. 1974. The Pleistocene changes of
vegetation and climate in tropical South America.
Journal of Biogeography 1: 3–26.
———. 1979. History of the flora, vegetation and climate
in the Colombian Cordillera Oriental during the last
five million years. Pp. 25–32 in Tropical botany, eds. K.
Larsen, B. Holm-Nielsen. London: Academic Press.
, W. 1821. Clinanthes lutea. Edwards’s Botanical
Register, Appendix 40.
. 1988. Clustal: a package
for performing multiple sequence alignment on a mi-
crocomputer. Gene 73: 237–244.
, C., S. W. L. J
. 1998. A molecular phylogeny of the subfam-
ily Arundinoidae (Poaceae) based on sequences of
rDNA. Austalian Systematic Botany 11: 41–52.
, A. T. 1985. Estudios sobre Amaryllidaceae. VI.
Sobre la inexistencia en Argentina del genero Phycella.
Lorentzia 5: 13–15.
, J. 1934. Familes of flowering plants, volume 2.
Monocot yledons, 1st edition. London: MacMillan and
———. 1959. Familes of flowering plants, volume 2. Mono-
cotyledons, 2nd edition. Oxford: Clarendon Press.
, M., A. K
, and S. K
. 1999. Phylogeny of Amaryllidaceae based on
matK sequence data. Japanese Journal of Plant Re-
search 112: 207–216.
. 1997. Phylogeny
and phylogeography of the circumpolar genus Frax-
inus (Oleaceae) based on internal transcribed spacer
sequences of nuclear ribosomal DNA. Molecular Phy-
logenetics and Evolution 7: 241–251.
. 1995. Molecular phylogeny of the
Papilionoideae (Family Leguminosae): rbcL gene se-
quences versus chemical taxonomy. Botanica Acta
108: 63–168.
, K. J. and R. K. J
. 1994. Comparisons of phy-
logenetic hypothese among different data sets in
dwarf dandelions (Krigia): additonal information
from internal transcribed spacer sequences of nuclear
ribosomal DNA. Plant Systematics and Evolution.
190: 157–185.
, M. 1995. Tribe Robinieae and allies: model groups
for assessing early tertiary northern latitude diversi-
fication of tropical legumes. Pp. 141–160 in Advances
in legume systematics 7: phylogeny, eds. M. Crisp and J.
J. Doyle. Kew: Royal Botanic Gardens.
——— and M. L
. 1993. Origins and relationships
of tropical North America in the context of the bor-
eotropics hypothesis. American Journal of Botany 80:
, C. 1753. Species plantarum. Stockholm.
, M. D., M. B. C
and M. W. C
. 1998. Systematics of Plumbagina-
ceae based upon cladistic analysis of rbcL sequence
data. Systematic Botany 23, 21–29.
, P. S. 1993. Cladistic analysis of sequences from
the internal transcribed spacers (ITS) of nuclear ri-
bosomal DNA of Nothofagus. American Journal of Bot-
any 80(6, supplement): 163.
, L. 1990. Hybrids and phylogenetic systematics.
I. Patterns of character expression in hybrids and their
implications for cladistic analysis. Evolution 44: 1685–
———. 1992. Hybrids and phylogenetic systematics. II.
The impcat of hybrids on cladistic analysis. Evolution
46: 1329–1346.
———. 1995. Hybridization and phylogenetics. Pp. 305–
331 in Experimental and molecular approaches to plant
biosystematics, eds. P. C. Hoch and A. G. Stephenson.
Saint Louis: Missouri Botanical Garden.
, A. W. 1987. A monograph of Eucrosia (Amaryl-
lidaceae). Systematic Botany 12: 429–461.
———. 1989. A monograph of the Amazon lilies, Eucharis
and Caliphruria (Amaryllidaceae). Annals of the Mis-
souri Botanical Garden 76: 136–220.
———. 1995. Towards a phylogeny of the Amaryllidaceae.
Pp. 169–179 in Monocotyledons: systematics and evolu-
tion, volume 1, eds. P. J. Rudall, P. J. Cribb, D. F. Cutler,
and C. J. Humphries. Kew: Royal Botanic Gardens.
——— and B. D
. 1985. A new species and a new
combination in the genus Eucrosia (Amaryllidaceae).
Brittonia 37: 47–55.
. 1998. Amaryllidaceae. Pp. 83–
110 in Families and genera of vascular plants, volume
3, ed. K. Kubitzki. Berlin: Springer-Verlag.
———, M. F. F
, Q-B. LI, F. Q. Z
M. W. C
. 1999. Systematics of Amaryllidaceae
based on cladistic analysis of plastid rbcL and trnL-F
sequence data. American Journal of Botany 86: 1325–
———, M. W. C
, Q-B. LI, D. S
S-Y. Y
. 2000. Phylogeny of Amaryllidaceae: mol-
ecules and morphology. Pp. 368–382 in Monocots, sys-
tematics and evolution, eds. K. Wilson and D. Morrison.
Sydney: CSIRO Press.
, M. and Q. C. B. C
. 1997. Phylogeny and
disjunct distribution: evolution of Saintpaulia (Gesner-
iaceae). Proceedings of the Royal Society of London,
Series B, Biological Sciences 264: 1827–1836.
. 1996. Tribes
and subtribes and some species combinations in
Amaryllidaceae J. St.-Hil. emend. R. Dahlgren et al.
1985. Feddes Repertorium. 107: S.c. 1–9.
, C. A. 1969. Cariotipos de nueve especies Ar-
gentinas de Rhodophiala, Hippeastrum, Zephyranthes y
Habranthus (Amaryllidaceae). Kurtziana 5: 67–87.
, C. A. and A. B. A
1975. El cariotipo
fundamental en el gene´ro Hippeastrum Herb. (Amar-
yllidaceae). Darwinia 19: 566–582.
, T. 1984. A cladistic analysis of
Haemanthus and Scadoxus. Nordic Jorunal of Botany 4:
, R. G. 1997. Is there a specific gene for a specific
problem? Advances i n pl ant molecular systematics. First
International Symposium of the Systematics Associ-
ation, University of Glasgow (Abstract). Glasgow:
Systematics Association.
, A. M., R. M. B
, and M. W. C
. 1997. Phylogenetic of sub-
tribe Orchidinae (Orchidoideae, Orchidaceae) based
on nuclear its sequences. 1. Intergeneric relationships
and polyphyly of Orchis sensu lato. Lindleyana 12:
. 1982. North Eastern
Brazil and the final separation of South America and
Africa. Palaeoecology 39: 71–85.
, P. H. and D. I. A
. 1974. Angiosperm bio-
geography and past continental movements. Annals
of the Misssouri Botanical Garden 61: 539–673.
, P. 1970. Rhodophiala cipoana Rav., sp. nov. Plant
Life 26: 86–87.
———. 1974. Studies in the genus Griffinia. Plant Life 30:
, K. E., C. S. K
, and B. A. M
. 1997. Mo-
lecular phylogenetic study of the ranunculaceae: util-
ity of the nuclear 26S ribosomal DNA in inferring
intrafamilial relationships. Molec. Phylogentics and
Evolution 8: 117–127.
, D. 1938. Karyotype alteration and phylogeny. IV.
Karyotypes in Amaryllidaceae with special reference
to the SAT-chromosomes. Cytologia 9: 203–242.
———. 1942. Karyotype alteration and phylogeny in Lil-
iaceae and allied families. Japanese Journal of Botany
12: 57–123.
. 1999. Molecular
phylogenetics of Fouquieriaceae: evidence f rom nucle-
ar rDNA ITS studies. American Journal of Botany 86:
, B. 1952. Chromosome counts of species and vari-
eties of garden plants. Annual report of the John In-
nes Horticultural Institute 42: 47.
, D. A. 1994. Systematics of Hessea, Strumaria and
Carpolyza (Amaryllideae: Amaryllidaceae). Contribu-
tions from the Bolus Herbarium 16: 1–162.
——— and H. P. L
. 1996. Phylogenetic relationships,
seed characters, and dispersal system evolution in
Amaryllideae (Amaryllidaceae). Annals of the Mis-
souri Botanical Garden 83: 362–386.
, M. D. 1996. TreeRot. Ann Arbor: University of
, Y., L. B. T
, and E. A. Z
. 1992. Nucleotide
sequences of the internal transcribed spacers and 5.8S
rRNA gene in Canella winterana (Magnoliales: Canel-
laceae). Nucleic Acids Research 20: 6101–6102.
, D. L. 1998. Phylogenetic analysis using parsi-
mony, v. 4.02 beta. Sunderland, MA: Sinauer Associ-
, D. W. 1988. Paleobiogeographic relationships of
the Paleogene flora from the southwestern USA: im-
plications for West Gondwanaland affinities. Palaeo-
geography, Palaeoclimatology and Palaeoecology 66:
———. 1990. Paleobiogeographic relationships of Angio-
sperms from the Cretaceous and early Tertiary of the
North American area. Botanical Review 56: 279–379.
———. 1995. Cretaceous to tertiary geologic and Angrios-
perm paleobiogeographic history of the Andes. Pp.
3–9 in Biodiversity and conservation of neotropical mon-
tane forests, eds. S. P. Churchill, H. Balslev, E. Forero,
and J. L. Luteyn. New York: New York Botanical Gar-
. 1997. The ClustalX windows
interface: flexible strategies for multiple sequence
alignment aided by quality analysis tools. Nucleic Ac-
ids Research 24: 4876–4882.
, B. H. 1985a. Perspectives on the origin of the
floristic similarity between eastern North America.
Journal of the Arnold Arboretum 66: 73–94.
———. 1985b. The Eocene North Atlantic land bridge: its
importance in Tertiary and modern phytogeography
of the northern hemispehere. Journal of the Arnold
Arborteum 66: 243–273.
, H. P. 1952. Biosystematic experiments involving
Zephyranthes, Habranthus and Amaryllis. Taxon 8: 121–
———. 1953. The genera Rhodophiala Presl., Phycella Lindl.
and Amaryllis L. Plant Life 9: 59–63.
———. 1963. Genera of the Amaryllidaceae. La Jolla:
American Plant Life Society.
——— and H. N. M
. 1949. Amaryllidaceae: Tribe
Amarylleae. La Jolla: Plant Life Society.
, J. W. 1997. When is a phylogenetic test good
enough? Memoirs de la Musee Nationale de Histoire
Naturelle 173: 31–45.
. 1990. Am-
plification and direct sequencing of fungal ribosomal
RNA genes for phylogenetics. Pp. 315–322 in PCR pro-
tocols: a guide to methods and applications, eds. M. Innis,
D. Gelfand, J. Sninsky and T. White. Orlando: Aca-
demic Press.
, J. A. 1975. Some aspects of plant geography of the
northern hemisphere during the late Cretaceous early
Tertiary. Annals of the Missouri Botanical Garden 62:
... The dispersal of the Old World genus Crinum L. (tribe Amaryllideae) into the Americas is considered a separate event (Meerow et al., 2003;Kwembeya et al., 2007). This Eurasian/American clade relationship has led some to declare it as evidence of a boreotropical (Wolfe, 1975;Tiffney, 1985a,b;Tiffney and Manchester, 2001) origin for the American clade (Christenhusz and Chase, 2013), despite cladistic patterns that are inconsistent with expectations of the boreotropics hypothesis, an absence of fossils (Meerow et al., 2000), and a lack of biogeographic analyses. ...
... The relationships of the endemic American genera were inferred using the internal transcribed spacer regions of nuclear ribosomal DNA (Meerow et al., 2000). These major relationships have also been supported by plastid data (Meerow et al., 1999(Meerow et al., , 2000Meerow and Snijman, 2005;Meerow, 2010). ...
... The relationships of the endemic American genera were inferred using the internal transcribed spacer regions of nuclear ribosomal DNA (Meerow et al., 2000). These major relationships have also been supported by plastid data (Meerow et al., 1999(Meerow et al., , 2000Meerow and Snijman, 2005;Meerow, 2010). The endemic American genera of the family were resolved as two major clades (Meerow et al., 1999(Meerow et al., , 2000. ...
Full-text available
One of the two major clades of the endemic American Amaryllidaceae subfam. Amaryllidoideae constitutes the tetraploid-derived (n = 23) Andean-centered tribes, most of which have 46 chromosomes. Despite progress in resolving phylogenetic relationships of the group with plastid and nrDNA, certain subclades were poorly resolved or weakly supported in those previous studies. Sequence capture using anchored hybrid enrichment was employed across 95 species of the clade along with five outgroups and generated sequences of 524 nuclear genes and a partial plastome. Maximum likelihood phylogenetic analyses were conducted on concatenated supermatrices, and coalescent-based species tree analyses were run on the gene trees, followed by hybridization network, age diversification and biogeographic analyses. The four tribes Clinantheae, Eucharideae, Eustephieae, and Hymenocallideae (sister to Clinantheae) are resolved in all analyses with > 90 and mostly 100% support, as are almost all genera within them. Nuclear gene supermatrix and species tree results were largely in concordance; however, some instances of cytonuclear discordance were evident. Hybridization network analysis identified significant reticulation in Clinanthus, Hymenocallis, Stenomesson and the subclade of Eucharideae comprising Eucharis, Caliphruria, and Urceolina. Our data support a previous treatment of the latter as a single genus, Urceolina, with the addition of Eucrosia dodsonii. Biogeographic analysis and penalized likelihood age estimation suggests an origin in the Cauca, Desert and Puna Neotropical bioprovinces for the complex in the mid-Oligocene, with more dispersals than vicariances in its history, but no extinctions. Hymenocallis represents the only instance of long-distance vicariance from the tropical Andean origin of its tribe Hymenocallideae. The absence of extinctions correlates with the lack of diversification rate shifts within the clade. The Eucharideae experienced a sudden lineage radiation ca. 10 Mya. We tie much of the divergences in the Andean-centered lineages to the rise of Frontiers in Plant Science | 1 November 2020 | Volume 11 | Article 582422 Meerow et al. Andean Amaryllidaceae the Andes, and suggest that the Amotape-Huancabamba Zone functioned as both a corridor (dispersal) and a barrier to migration (vicariance). Several taxonomic changes are made. This is the largest DNA sequence data set to be applied within Amaryllidaceae to date.
... Clinanthus Herbert (1821: 40) was segregated from Stenomesson Herbert (1821: 40) by Meerow et al. (2000), on the basis of nrDNA ITS sequences indicating that the latter was polyphyletic. Stenomesson, all of which have pseudopetiolate leaves, are now placed in the tribe Eucharideae (Meerow 2010, Meerow et al. 2000. ...
... Clinanthus Herbert (1821: 40) was segregated from Stenomesson Herbert (1821: 40) by Meerow et al. (2000), on the basis of nrDNA ITS sequences indicating that the latter was polyphyletic. Stenomesson, all of which have pseudopetiolate leaves, are now placed in the tribe Eucharideae (Meerow 2010, Meerow et al. 2000. Clinanthus is endemic to the Andes of South America and occurs from southern Ecuador through Peru, Bolivia and northern Chile in both coastal desert and mountain habitats, from moist to seasonally dry (Esquerre & Meerow 2020). ...
Clinanthus fabianae and C. thiagoi are described from Ancash and La Libertad departments of Peru. Both species have similar perigone coloration, broadly lobed staminal coronas, and incurved free filaments. They may represent a distinct species group, along with the previously described C. inflatus. The species can be separated by number of flowers, pedicel length, flower habit, tube length, morphology and pattern of dilation, and color of the tepal apiculum.
... Two species of the genus Pancratium L., also from GenBank, were used as outgroup. DNA extraction, amplification, and sequencing protocols were as described in Meerow et al. (2000Meerow et al. ( , 2006. Ten thousand simple addition replicates of parsimony Jackknife analysis were run using PAUP v. 4a169 (Swofford 2002), and the Jackknife 50% consensus tree was visualized using FigTree v. 1.4.4 (Rambaut 2019). ...
... Although of limited use in resolving phylogenetic relationships at the infrageneric level, nrITS proved to be sufficiently informative to place most American species of Amaryllidaceae in the genera and tribes in which they are currently circumscribed (Meerow et al. 2000;Meerow 2010;Oliveira 2012;García et al. 2014). Hippeastrum velloziflorum is robustly positioned as the first branch of subgenus ...
Full-text available
In 2015, Brazil faced the worst environmental disaster in its history, when the collapse of an iron ore dam dumped millions of tons of tailings into the Doce River. In this paper, we describe two Hippeastrum species native to localities directly involved in the tragedy. The dam was located in the foothills of Serra do Caraça, a mountain range in the state of Minas Gerais, from where we describe the endemic H.carassense; H.velloziflorum was first found on an inselberg located on the banks of the Doce River, in the neighboring state of Espírito Santo. Comments on their distribution, ecology, and phenology are provided, as well as comparisons with the most similar taxa. The conservation status of the two new species is preliminarily assessed, and both are considered threatened with extinction. We also compared their leaf anatomy and micromorphology with related species of Amaryllidaceae. Based on nrDNA ITS, we infer the phylogenetic position of H.velloziflorum, a taxon with several unique morphological characters for Hippeastrum, as the first branch in subgenus Hippeastrum. The placement of H.velloziflorum in Hippeastrum is also supported by anatomical and cytological data. The somatic chromosome number was 2n = 22, and the karyotype formula was 2n = 8m + 12sm + 2st chromosome pairs. An identification key to the species of Hippeastrum occurring in the Doce and Jequitinhonha River basins is presented.
... Zephyranthes fosteri (Fig. 1) and Z. alba ( Fig. 2) (Amaryllidaceae: Amaryllidoideae) belong to the Hippeastreae tribe, which contains 10-13 genera and 180 species. In America, several diversification centers have been identified in Chile, Argentina, Brazil (Meerow et al., 2000;Arroyo-Leuenberger and Dutilh, 2008;Arroyo-Leuenberger and Leuenberger, 2009) and Mexico (García et al., 2014). Z. fosteri (Fig. 1) is a perennial herb with a height between 5 and 45 cm with bulbs between 1.5 and 4.5 cm. ...
Zephyranthes (Amaryllidaceae) is a taxonomically complex genus due to the frequent overlap of interspecific morphological variation. In Mexico, Z. alba and Z. fosteri are herbaceous plants that, when distributed in sympatry, generate individuals with complex patterns of morphological variation, leading to taxonomic confusion. Therefore, it is necessary to first characterize these species in allopatric populations. In this contribution, molecular, morphological, and alkaloid profiles were used to characterize both species in allopatric sites. Our results show that Z. alba and Z. fosteri allopatric populations are two well-defined genetic and morphological groups. Flower-related characters were the ones that best allowed us to distinguish between species. In a similar fashion, the alkaloid profile showed remarkable differences among species: four alkaloids were specific to Z. alba and five to Z. fosteri. Lycorine (43.3 ̶ 88.8%) and galanthamine (87.7 ̶ 91.4%) were the most abundant alkaloids for each species, respectively. In conclusion, Z. fosteri and Z. alba exhibit noticeable differences when distributed in allopatry. In addition, Z. fosteri has greater genetic and phenotypic plasticity compared to Z. alba, which could be related to the former's ability to colonize new habitats. Finally, the molecular, genetic and chemical markers developed here will provide a framework to further studies aiming to explore if hybridization among Z. alba and Z. fosteri occurs in sympatric populations.
... Nic.García, formally known as Rhodophiala phycelloides (Herb.) Hunz belongs to the highly polyphyletic family Amaryllidaceae J. St.-Hil., a group of monocotyledonous, geophytic, bulbous, petaloid, cosmopolitan plants (Meerow et al., 2000). The family is composed of 1600 species of approximately 75 genera and is widely distributed in South America, the Mediterranean and South Africa (Xu & Chang, 2017). ...
Full-text available
Sporadic rains in the Atacama Desert reveal a high biodiversity of plant species that only occur there. One of these rare species is the ''Red añañuca" (Zephyranthes phycelloides), formerly known as Rhodophiala phycelloides. Many species of Zephyranthes in the Atacama Desert are dangerously threatened, due to massive extraction of bulbs and cutting of flowers. Therefore, studies of the biodiversity of these endemic species, which are essential for their conservation, should be conducted sooner rather than later. There are some chloroplast genomes available for Amaryllidaceae species, however there is no complete chloroplast genome available for any of the species of Zephyranthes subgenus Myostemma. The aim of the present work was to characterize and analyze the chloroplast of Z. phycelloides by NGS sequencing. The chloroplast genome of the Z. phycelloides consists of 158,107 bp, with typical quadripartite structures: a large single copy (LSC, 86,129 bp), a small single copy (SSC, 18,352 bp), and two inverted repeats (IR, 26,813 bp). One hundred thirty-seven genes were identified: 87 coding genes, 8 rRNA, 38 tRNA and 4 pseudogenes. The number of SSRs was 64 in Z. phycelloides and a total of 43 repeats were detected. The phylogenetic analysis of Z. phycelloides shows a distinct subclade with respect to Z. mesochloa. The average nucleotide variability (Pi) between Z. phycelloides and Z. mesochloa was of 0.02000, and seven loci with high variability were identified: psbA, trnS GCU-trnG UCC , trnD GUC-trnY GUA , trnL UAA-trnF GAA , rbcL, psbE-petL and ndhG-ndhI. The differences between the species are furthermore confirmed by the high amount of SNPs between these two species. Here, we report for the first time the complete cp genome of one species of the Zephyranthes subgenus Myostemma, which can be used for phylogenetic and population genomic studies.
... Sequences were aligned and adjusted using ContigExpress (Lu and Moriyama 2004), then assembled and checked manually. Newly generated DNA sequence data of three individuals of S. yunnanensis from the typical locality were integrated with data of three cpDNA regions (ndhF, matK, rbcL) and nrITS of 33 taxa in subfamily Amaryllidoideae of Amaryllidaceae selected from previous datasets (Ito et al. 1999, Meerow et al. 1999, 2000 and NCBI, including all eleven currently accepted genera in the Eurasian clade. One species from Amaryllidaceae subfamily Allioideae was was included as the outgroup. ...
Shoubiaonia yunnanensis is described here as a new genus and species of Amaryllidaceae from Yunnan Province, southwest China. Phylogenetic analyses based on DNA sequences of nuclear ribosomal ITS and three plastid regions (matK, ndhF and rbcL) strongly support Shoubiaonia as a member in the tribe Lycorideae of the Eurasian clade of Amaryllidaceae. However, morphologically S. yunnanensis is readily distinguished from the other genera in the Eurasian clade by its 2-valved spathe, 3-lobed stigma, ovary with two ovules per locule and silver black, subglobose seeds. The new genus is the second genus of the subfamily Amaryllidoideae distributed in natural habitat in China.
... La familia Amaryllidaceae J. St.-Hil. tiene una distribución cosmopolita, aunque predomina en el trópico (Meerow et al., 1999(Meerow et al., , 2000; y con una taxonomía aún compleja (Oliveira, 2012). En 2003 Amaryllidaceae, Alliaceae e Agapantaceae fueron agrupadas en una sola familia (The Angiosperm Phylogeny Group, 2003) que representa 39 géneros en la región neotropical. ...
Full-text available
Hippeastrum angustifolium Pax es una hierba bulbosa de la familia Amaryllidaceae, distribuida en los humedales de Argentina, Paraguay, Uruguay y sur-suroeste de Brasil. En este trabajo se compiló información sobre la distribución de esta especie en Paraguay. Se encontraron 15 registros oficiales para Paraguay, de los cuales, solamente 2 especímenes fueron colectados en las últimas dos décadas. La información disponible sobre la distribución y el estado de conservación de la especie fue escasa.
... (García et al., 2014(García et al., , 2019. Tetraploid Andean Amaryllidaceae are presented by Eustephieae, Hymenocallideae, Clinantheae, and Eucharideae tribes (Meerow et al., 2000). ...
... Rhodophiala C. Presl, Amaryllidaceae, is a plant genus that phylogenetically belongs to the tribe Hippeastreae, subtribe Hippeastrinae (Meerow et al. 2000). Rhodophiala plants have a tunicate bulb 4-6 cm in diameter, which is set 20-30 cm underground (Olate and Bridgen 2005). ...
Rhodophiala bifida (Herb.) Traub, Amaryllidaceae, is a species native to South America known to produce alkaloids with potential pharmacological uses such as montanine, which has anti-inflammatory potential. R. bifida could be applied as a natural source of montanine. It is important to understand the genetic diversity of this species in order to assess the sustainable use of this plant. The aim of this study was to evaluate the genetic diversity and chemical profiles of the two known natural populations of R. bifida in Brazil. This report is the first population genetic study of R. bifida. We studied 93 individuals with six Inter-Simple Sequence Repeats (ISSR) primers. A total of 79 loci were amplified. Our results showed high-population structure (Fst = 0.16), with greatest genetic variation at the intrapopulation level. Genetic analyses separated the individuals of R. bifida into two clusters that corresponded to each of the natural populations. Chemical profile evaluation was carried out on dried bulbs, leaves, flowers, and flower scapes by liquid chromatography and mass spectrometry. Montanine and nangustine were the main metabolites identified in both populations. These alkaloids concentrations differed by population and by plant part.
Zephyranthes citrina is an ornamental American bulbous plant used as an ornamental garden crop for the aesthetic qualities of its yellow perigonium. The objective of this work was to characterize the species by classical chromosome staining and fluorochrome banding. A sporophytic chromosome number of 2n=8x=48 chromosomes was observed, being the karyotypic formula 20 m + 26 sm + 2 st. Satellites were detected in the short arm of metacentric chromosomes 8, 9, 11 and 12, which colocalized with constitutive heterochromatin CMA+/DAPI-/0 bands. The karyotype comprised chromosome pairs with terminal constitutive heterochromatin bands that included satellites and heteromorphic clusters indicating that it is an allooctoploid. These results will be used as a tool for monitoring genetic improvement, in interspecific crosses and its progenies and in biotechnological procedures by in vitro culture. Key words: constitutive heterochromatin, chromosome banding, bulbous, plant genetic resources, karyotype
The vegetation of northern Latin America is a complex of assemblages arriving by different routes and at various times, depending upon the physical conditions and climates prevailing along the North Atlantic and Panamanian land bridges. The result is a modern lowland and lower montane vegetation in northern Latin America that includes four biogeographic components: 1) plants of Gondwanan; 2) Laurasian ancestry that arrived via the North Atlantic and, to a lesser extent, across Beringia, during the late Mesozoic and Paleogene; 3) ancient Gondwanan elements established in South America before separation from Africa, and moving into northern Latin America during the late Mesozoic and Paleogene; and 4) Gondwanan (South American) introductions across the isthmus in Neogene times. -from Author
Among the infrafamilial ranks of Amaryllidaceae J.St.-Hil. emend. R.Dahlgren and al. = Amaryllidaceae s. str. the level of tribes is the most important one. At this level the taxonomic groups form clear-cut entities which have in most cases a common distribution on one continent or a portion of it. In nearly complete agreement with Dahlgren et al. 1985 we distinguish ten tribes (instead of nine) of which five are identical. Three further tribes have now the same names but a different circumscription: for the two genera Gethyllis and Apodolirion, discussed by Dahlgren et al. 1985 as 'possibly related to the Haemantheae', the tribe Gethyllidae is validated here; our Narcisseae s.l. include Galantheae; the two Australian genera Calostemma and Proiphys (radiating to southern Asia) are described as a new tribe. The rank of subfamily is not used. The authorship of five tribes accepted by us has changed as compared with Traub 1963, 1970 by taking into account earlier publications. Nineteen subtribes are recognized, nine of which had tribal level in the system of Traub. Further seven new subtribes are proposed here.
Total DNA was extracted from 67 species (30 genera) of the subfamily Papilionoideae (family Leguminosae). The rbcL gene was amplified by polymerase chain reaction (PCR) and sequenced directly. RbcL sequences were evaluated with character-state (maximum parsimony; PAUP) and distance methods (neighbour-joining; MEGA). Morphology-based classifications of tribes and subtribes are mostly congruent with rbcL phylogeny. Differences occur for members of the genus Sophora and for intratribal relationships within the Genista/Cytisus complex: Sophora appears in two clades; one clade with S. japonica at the base of the Papilionoideae and a second more advanced group (with S. davidii, S. jaubertii, and S.flavescens) which represents a sister clade of the Thermopsideae. The Cytisus complex includes the genera Cytisus, Chamaecytisus, Calicotome and Spartocytisus but not Cytisophyllum which appears as a distinct member of the Genista complex. Chamaespartium sagittale is very close to Genista supporting the view that it does not represent an independent genus but should be treated as Genista sagittalis. Close to Genista are Teline and Spartium, whereas Argyrolobium, Retama, Cytisophyllum, Ulex, Petteria, Adenocarpus, Chamaespartium tridentatum and Laburnum can be considered as “outliers” of the Genista-group sensu stricto. RbcL phylogeny is compared with profiles of alkaloids and other natural products. First results indicate that secondary metabolite profiles, if compared with morphology or rbcL sequences, are of limited value as a taxonomic marker.
At the end of the Cretaceous there was a possibility for relatively direct floristic interchange between South America and tropical North America via island hopping along the proto- Antilles. Uplift of the Andes, mostly in Neogene time, led to an incredible burst of speciation in a number of Gondwanan families. A similar evolutionary explosion in the same taxa also took place in Costa Rica and Panama. The taxonomic groups that have undergone this evolutionary explosion have distributional centers in the N Andean region and S Central America, are poorly represented in Amazonia, and consist mostly of epiphytes, shrubs, and palmettos; their pollination systems suggest that coevolutionary relationships with hummingbirds, nectar-feeding bats, and perhaps such specialized bees as euglossines, have played a prominent role in their evolution. The evolutionary phenomena associated with the Andean uplift account for almost half of the total Neotropical flora and are thus largely responsible for the excess floristic richness of the Neotropics. Closing of the Panamanian isthmus in the Pliocene led to 1) southward migration of some Laurasian taxa into the Andes where they have become ecologically dominant despite undergoing little speciation, at least in woody taxa, and 2) northward invasion of lowland Gondwanan taxa of canopy trees and lianas into Central America, leading to their ecological dominance in lowland tropical forests throughout the region, despite little significant speciation in Central America.-from Author
Amino acid sequence data are available for ribulose biphosphate carboxylase, plastocyanin, cytochrome c, and ferredoxin for a number of angiosperm families. Cladistic analysis of the data, including evaluation of all equally or almost equally parsimonious cladograms, shows that much homoplasy (parallelisms and reversals) is present and that few or no well supported monophyletic groups of families can be demonstrated. In one analysis of nine angiosperm families and 40 variable amino acid positions from three proteins, the most parsimonious cladograms were 151 steps long and contained 63 parallelisms and reversals (consistency index = 0.583). In another analysis of six families and 53 variable amino acid positions from four proteins, the most parsimonious cladogram was 161 steps long and contained 50 parallelisms and reversals (consistency index = 0.689). Single changes in both data matrices could yield most parsimonious cladograms with quite different topologies and without common monophyletic groups. Presently, amino acid sequence data are not comprehensive enough for phylogenetic reconstruction among angiosperms. More informative positions are needed, either from sequencing longer parts of the proteins or from sequencing more proteins from the same taxa.