全 文 ：Journal of Systematics and Evolution 46 (3): 341–348 (2008) doi: 10.3724/SP.J.1002.2008.08025
(formerly Acta Phytotaxonomica Sinica) http://www.plantsystematics.com
Phylogeny of Catalpa (Bignoniaceae) inferred from sequences of
chloroplast ndhF and nuclear ribosomal DNA
Jianhua LI*
(Arnold Arboretum, Harvard University Herbaria, 22 Divinity Avenue, Cambridge, MA 02138, USA; Adjunct Faculty of College of Life Sciences,
Zhejiang University, Hangzhou 310029, China)
Abstract Phylogenetics of Chilopsis and Catalpa (Bignoniaceae) was elucidated based on sequences of chloro-
plast ndhF and the nrDNA ITS region. In Bignoniaceae, Chilopsis and Catalpa are most closely related as sister
genera. Our data supported section Macrocatalpa of the West Indies and section Catalpa of eastern Asian and
North American continents. Within section Catalpa, Catalpa ovata of eastern Asia form a clade with North
American species, C. speciosa and C. bignonioides, while the other eastern Asian species comprise a clade where
C. duclouxii is sister to the clade of C. bungei and C. fargesii. The Caribbean species of Catalpa diverged early
from the continental species. More studies are needed to test whether the phylogenetic pattern is common in
eastern Asian-North American disjunct genera with species in the West Indies.
Key words Bignoniaceae, Catalpa, eastern Asia-North America, ndhF gene, nrDNA ITS sequences, phylogeny,
West Indies.
Catalpa Scop. (Bignoniaceae), an intercontinen-
tal disjunct genus, consists of ten species, with two
species in eastern North America (ENA), four in
eastern Asia (EAS), and four in the West Indies (WI)
(Li, 1952; Paclt, 1952; Gentry, 1992). Species of
Catalpa are semievergreen or deciduous trees with
opposite or whorled leaves. Their bisexual flowers are
arranged in a raceme or panicle, and have two fertile
stamens, a 2-lipped corolla, and tetrad pollen grains
(Gentry, 1992). Four semievergreen species of section
Macrocatalpa Griseb. are distributed in the WI,
including C. brevipes Urban, C. longissima Sims, C.
macrocarpa Ekman, and C. purpurea Griseb. (Man-
ning, 2000). Species of section Catalpa are deciduous
and disjunctly distributed between EAS and ENA.
Catalpa bignonioides Walter, or the southern catalpa,
has an original distribution in northern Florida and
southwestern Georgia to southern Alabama and
eastern Mississippi (Duncan & Duncan, 1988), Lou-
isiana (Little, 1979), or easternmost Texas (Weniger,
1996). Catalpa speciosa Warder ex Engelm., the
western catalpa, is native to the Mississippi River
drainage basin from central Illinois and Indiana to
northeastern Arkansas, western Tennessee (Beillman
1946; Little, 1979), and Louisiana (Burk & McMaster,
1988; Duncan & Duncan, 1988; Thomas & Allen,
1996). Both species have spread to other areas as a
result of their cultivation as garden or lawn trees
(Beillmann, 1946). Catalpa ovata G. Don is distrib-
uted in central and northern China. Its young leaves
are edible, while the extract of mature leaves and
barks has been used as pesticide and in traditional
Chinese medicine (Wang, 1990). Catalpa bungei C.
A. Mey. and C. fargesii Bureau are distributed in
central to southwestern China, and the latter has a
glabrous form, namely, C. fargesii f. duclouxii (Don)
Gilmour (Gilmour, 1936; =C. duclouxii Dode). Ca-
talpa tibetica Forrest is endemic to southwestern
China and shares creamy-yellow flowers with C.
ovata (Forrest, 1921).
Catalpa shares many reproductive characters
with Chilopsis D. Don, a monotypic genus of the
Chihuahuan and Sonoran deserts of northern Mexico
and the southwestern United States (Henrickson,
1985; Gentry, 1992). Sterile hybrids have also been
formed between Catalpa species and Chilopsis lin-
earis Sweet (Rusanov, 1964; Li et al., 2006). There-
fore, Catalpa has been considered as most closely
related to Chilopsis in Bignoniaceae (Gentry, 1992).
In this study, interspecific relationships of Ca-
talpa were estimated based on sequences of chloro-
plast gene ndhF and the internal transcribed spacers of
nuclear ribosomal DNA (nrDNA ITS). Both DNA
regions have been widely used for resolving lower
level relationships of plant groups (Baldwin et al.,
1995; Soltis et al., 1998; Davis et al., 2002). Specifi-
cally, I focused on the following three questions: (1)
Are Chilopsis and Catalpa most closely related genera
in Bignoniaceae? (2) Are sections Macrocatalpa and
Catalpa each monophyletic? (3) Are North American

———————————
Received: 26 February 2008 Accepted: 20 March 2008
* E-mail: jli@oeb.harvard.edu; Tel.: 617-496-6429; Fax: 617-495-9484.
Journal of Systematics and Evolution Vol. 46 No. 3 2008 342
species more closely related to the West Indian spe-
cies or eastern Asian species?
1 Material and Methods
1.1 Taxon sampling
Thirteen samples of Catalpa and Chilopsis were
included representing all species of the two genera
except for Catalpa brevipes, C. purpurea, and C.
tibetica, whose materials were unavailable for mo-
lecular study. Previous studies have recognized the
monophyly of Bignoniaceae (Spangler & Olmstead,
1999; Olmstead et al., 2000). Nevertheless, the phy-
logenetic relationships of Chilopsis and Catalpa with
other genera of the family remain unclear. Therefore, I
also included 18 other bignoniaceous genera repre-
senting all tribes of the Bignoniaceae (Spangler &
Olmstead, 1999) and their ndhF sequences were
obtained from the GenBank (Table 1).
1.2 Molecular techniques
DNAs were extracted from silica-gel dried or
fresh leaf tissue using a Qiagen DNeasy Plant Mini
Kit (cat. # 69104, Germantown, MD). The nrDNA
ITS region was amplified using primers ITS4 (White
et al., 1990) and ITSLeu (Baum et al., 1998). A 25 µL
PCR reaction included 2.5 µL Taq polymerase buffer
(10 ×), 4 µL of dNTP (2.5 mmol/L), 2 µL of MgCl2
(25 mmol/L), 1 µL of each primer (10 µmol/L), 0.2
µL of Taq polymerase (5 U/µL), 50–100 ng DNA, 2
µL of DMSO (dimethyl sulfoxide), and an appropriate
amount of sterilized water. The thermocycler program
consisted of the following steps: a hot-start of 94 ℃
for 3 min, 35 cycles of 94 ℃ for 1 min, 55 ℃ for 2
min, and 72 ℃ for 1 min. The final cycle was fol-
lowed by an additional 7 min extension at 72 ℃. The
amplified products were purified using a Qiagen Gel
Purification Kit (Santa Clarita, CA).
The 3′ end segment of the ndhF gene was ampli-
fied in an MJ-PT200 Thermocycler using primers
ndhF972F and 2210R (Olmstead & Sweere, 1994).
A 25 µL reaction contained 50–100 ng of genomic
DNA, 4 µL of DNTPs (2.5 mmol/L), 3 µL of MgCl2,
2.5 µL of Taq polymerase buffer (10 ×), 0.3 µL of Taq
polymerase (5 U/µL), 1 µL of each primer (10
µmol/L), and sterilized water. The PCR program
consisted of a 3 min hot-start at 94 ℃ and 35 cycles
of 1 min denature at 94 ℃, 1.5 min annealing at 50
℃, and 2 min extension at 72 ℃. The cycles were
followed by an additional 7 min extension at 72 ℃.
Our initial attempt to do direct sequencing of the PCR
products failed, indicating that there might be se-
quence polymorphisms in species of Catalpa and
Chilopsis. Therefore, the PCR products were cloned
using a standard pGEM T Easy Vector System
(Promega, Madison). From the same species more
than one accession was sampled to detect for intras-
pecific variation and for each accession three to eight
clones were sequenced. Repeated DNA extraction,
PCR, and sequencing reactions were conducted to
detect PCR errors that may have led to sequence
variation. Clones and PCR products were sequenced
using the Dideoxy Terminator Chemistry with an ABI
BigDye Cycle Sequencing Ready Reaction kit. Se-
quences were analyzed using an ABI 3100 or 3730
Genetic Analyzer, and were edited using Sequencher
(Version 4.1, GeneCode Inc., Ann Arbor, MI).
1.3 Phylogenetic analyses
In all phylogenetic analyses, characters were
weighted equally and their state changes were treated
as unordered. Gaps in sequences were treated as
missing data. Sequences of both ndhF and nrDNA ITS
regions were aligned readily by sight. Phylogenetic
analyses were conducted using Maximum parsimony
(MP) and maximum likelihood (ML) methods in
PAUP* (Swofford, 2002). Heuristic tree search in MP
analyses included the following options: 1000 repli-
cates of random sequence addition with 1 tree held
each replicate, TBR branch swapping, Multrees on,
and steepest descent off. Branch and bound tree search
was conducted using default options in PAUP*.
Bootstrap analyses of 1000 replicates (Felsenstein,
1985) were carried out to estimate support for indi-
vidual clades using options as in parsimony analyses
with heuristic tree search except for simple sequence
addition. Modeltest (version 3.06, Posada & Crandall,
1998) was used to select the best model for molecular
data sets. Then, the estimated parameters of the se-
lected model were applied in the ML analyses. The
congruence of the nrDNA ITS and ndhF data sets was
evaluated by comparing tree topologies from individ-
ual data sets to check whether there are well-supported
(bootstrap support, bs>70%) but conflicting clades.
2 Results
2.1 Sequence characteristics
The 3′ portion of the ndhF gene and the entire
nrDNA ITS region were newly obtained from the
thirteen accessions of Chilopsis and Catalpa and the
sequences have been submitted to GenBank (Table 1).
The segment of the ndhF gene corresponded to the
region between positions 1011–2096 of the ndhF gene
in Campsis radicans (AF102626) and was 1086 base
LI: Catalpa phylogenetics

343
Table 1 Species included in the study
GenBank Accessions Species Source and voucher
ndhF nrDNA ITS
Arrabidaea pubescens (L.) A. H. Gentry – AF102625 –
Amphitecna apiculata A. H. Gentry – AF102624 –
Campsis radicans (L.) Seem. – AF102626 –
Crescentia portoricensis Britton – AF10262 –
Cybistax donnell-smithii (Rose) Seibert – AF102628 –
Cydista aequinoctialis Miers – AF102629 –
Eccremocarpus scaber Ruiz & Pav. – AF102630 –
Jacaranda sparrei A. H. Gentry – AF102631 –
Kigelia africana (Lam.) Benth. – AF102632 –
Macfadyena unguis-cati (L.) A. H. Gentry – AF102633 –
Martinella obovata Bureau & K. Schum. – MAXCPNDH –
Ophiocolea floribunda (Boj. ex Lindl.) H. Perrier – AF102634 –
Oroxylum indicum (L.) Benth. ex Kurz – AF102635 –
Podranea ricasoliana Sprague – AF102637 –
Radermachera frondosa Chun & F. C. How – AF102638 –
Tabebuia heterophylla (DC.) Britton – TABCPNDH –
Pandorea jasminoides Schum. – AF102636 –
Tecoma stans Juss. – AF130145 –
Chilopsis linearis Sweet Genhua Niu, Texas A & M Univ., jli4196 DQ411419 AY178657
Catalpa longissima Sims FBG 961378*A, jli3181 DQ411414 AY486294
Catalpa macrocarpa Ekman FBG 66346*A, jli3182 DQ411415 AY486295
Catalpa duclouxii Dode 1 AA 642-94A, jli3233 DQ411418 AY486296
Catalpa duclouxii 2 FM Jun Wen 5705, jli3250 DQ411417 AY486297
Catalpa bungei C. A. Mey. AA 12927A, jli3221 DQ411410 AY486299
Catalpa fargesii Bureau AA 12222B, jli3211 DQ411411 AY486300
Catalpa ovata G. Don 1 AA 516-87A, jli3212 DQ411412 AY486302
Catalpa ovata 2 AA 98-61A, jli3214 DQ411413 AY486303
Catalpa speciosa Warder ex Engelm. 1 AA 1245-79C, jli3210 DQ411408 AY486305
Catalpa speciosa 2 AA 927-58B, jli3209 DQ411409 AY486306
Catalpa speciosa 3 AA 131-54A, jli3230 DQ411407 AY486307
Catalpa bignonioides Walter AA 592-60C, jli3237 DQ411416 AY486308
AA, Arnold Arboretum accessions with wild provenance; FBG, Fairchild Botanical Gardens; FM, Field Museum at Chicago; sequences newly obtained
are highlighted in bold.


pairs (bp) in all genera of Bignoniaceae. There were
305 variable sites, 115 of which were parsimony
informative (10.6%). Some ndhF clones of Catalpa
and Chilopsis were pseudogenes as indicated by
frameshift indels and base mutations, and premature
codons (Li et al., unpublished data). However, such
pseudogenes were not available from Catalpa longis-
sima or C. macrocarpa. Thus, they were excluded
from phylogenetic analysis.
Sequence heterogeneity of the nrDNA ITS region
within species was not observed in any of the taxa
sampled. Sequence length ranged from 617–631 bp in
Catalpa and Chilopsis. The GC content was ca. 61%
and did not differ significantly, as judged by the
Chi-square test of homogeneity of base frequencies
across the taxa in PAUP*. Sequence divergence
ranged from 7.4%–8.7% between Chilopsis linearis
and Catalpa and from 6.4%–8.3% between sections
Macrocatalpa and Catalpa. Within sections sequence
divergence ranged from 0.2%–5.8%. The sequence
alignment of nrDNA ITS region generated a data set
of 634 sites, requiring 9 single base gaps. There were
95 variable sites, 66 of which were parsimony infor-
mative.
Journal of Systematics and Evolution Vol. 46 No. 3 2008 344
2.2 Phylogenetic relationships
2.2.1 ndhF data Parsimony analyses of ndhF gene
sequence data generated 14 trees in one island of 441
steps. Figure 1 is the strict consensus rooted with
Jacaranda sparrei A. H. Gentry and Podranea ri-
casoliana Sprague (CI=0.82, RI=0.79). There were six
major branches: (1) Eccremocarpus scaber Ruiz &
Pav.; (2) Pandorea jasminoides Schum., Tecoma stans
Juss., and Campsis radicans (L.) Seem.; (3) Amphi-
tecna apiculata A. H. Gentry, Crescentia portoricen-
sis Britton, Cybistax donnell-smithii (Rose) Seibert,
Tabebuia heterophylla (DC.) Britton, Kigelia africana
(Lam.) Benth., Ophiocolea floribunda (Boj. ex Lindl.)
H. Perrier, and Radermachera frondosa Chun & F. C.
How (bs=60%); (4) Arrabidaea pubescens (L.) A. H.
Gentry, Martinella obovata Bureau & K. Schum.,
Macfadyena unguis-cati (L.) A. H. Gentry, and Cy-
dista aequinoctialis Miers (bs=100%); (5) Oroxylum
indicum (L.) Benth. ex Kurz; and (6) Chilopsis and
Catalpa (bs=96%). However, relationships among
these clades were not resolved. Chilopsis was sister to
Catalpa, whose species formed a strongly supported
clade (bs=100%). Within Catalpa, there were two
clades corresponding to sections Macrocatalpa (bs=
100%) and Catalpa (bs=94%). Section Catalpa had
three branches whose relationships were not resolved.
North American species, C. speciosa and C. bignoni-
oides, formed a clade (bs=99%). Three Asian species
also formed a clade including C. duclouxii, C. bungei,
and C. fargesii. However, C. ovata, another Asian
species, did not cluster with either of the previous
clades.
Modeltest selected GTR+G as the best model of
evolution for ndhF sequences and the estimated
parameters were as follows: base frequencies (A=
0.3079, C=0.1459, G=0.1697, T=0.3764), rate matrix
(A-C=2.0086, A-G=2.6265, A-T=0.1926, C-G=
2.467, C-T=2.6265, G-T=1), and Gamma shape
parameter=0.8248. ML analyses using the estimated
parameters produced a single tree with a likelihood of
–ln=4057.8275. The ML tree topology (not shown)
suggested the same relationships of Chilopsis and
Catalpa as in the parsimony tree (Fig. 1).
2.2.2 nrDNA ITS region The ITS data set con-
tained 13 samples of Chilopsis and Catalpa because
our ndhF data strongly support the sister relationship
of the two genera (see above). Furthermore, this
relationship has long been suggested based on mor-
phological evidence (Gentry, 1992). Parsimony
analyses of the 13-taxon data set using branch and
bound tree search produced 3 trees of 117 steps, one
of which is shown in Fig. 2a (CI=0.88, RI=0.92).
Sections Macrocatalpa and Catalpa each formed their
own clades (bs=100% and 92%, respectively). Within
section Catalpa there were two clades, including (C.
ovata, (C. speciosa, C. bignonioides)) and (C. du-
clouxii, (C. bungei, C. fargesii)). All clades were
strongly supported. ML analyses were performed
using the HKY+G model selected by Modeltest and
the estimated parameters included base frequencies
(A=0.1915, C=0.3105, G=0.2899, T=0.2082), Ti/Tv
ratio=2.57, and Gamma shape parameter=0.1971. The
ML tree with a likelihood of –ln=1513.9849 was
entirely congruent with Fig. 2a except that three
accessions of C. speciosa formed a moderately sup-
ported clade (bs=75%, Fig. 2b).
2.2.3 ndhF+nrDNA ITS Phylogenetic trees
inferred from the ndhF gene (Fig. 1) and nrDNA ITS
(Fig. 2) were congruent. Therefore, the two data sets
were combined resulting in a matrix of 1720 sites.
Parsimony analyses using branch and bound tree
search generated 3 tree of 173 steps (CI=0.91,
RI=0.93). These trees differed in the relationships of
the three accessions of C. speciosa to C. bignonioides.
The consensus tree (not shown) was congruent with
the ITS tree (Fig. 2) with higher bootstrap support for
all clades.
3 Discussion
Catalpa and Chilopsis are nearly identical in
fruit, seed, embryo, style, and anther characteristics
(Gentry, 1980; Henrickson, 1985; Manning, 2000).
They share pollen tetrads with sculpturing limited to
coarsely reticulate areoles, a unique pollen type in the
Bignoniaceae (Gentry & Tomb, 1979). In addition, an
intergeneric sterile hybrid has been reported between
Catalpa and Chilopsis (Rusanov, 1964) and has
recently been confirmed based on molecular data (Li
et al., 2006). Chilopsis is sister to the clade containing
all species of Catalpa, offering strong support for the
close affinity of the two genera (Fig. 1).
Catalpa has been divided into sections Macro-
catalpa and Catalpa (Paclt, 1952; Gentry 1980, 1992;
Manning, 2000). In the nrDNA ITS, chloroplast ndhF
gene, and combined trees (Figs. 1, 2), the two species
of section Macrocatalpa form a robust clade, so do
species of section Catalpa. Their monophyly also gets
support from morphology. For example, leaves are
elliptic and evergreen in section Macrocatalpa (vs.
broadly ovate and deciduous in section Catalpa)
(Paclt, 1952). Foliar nectaries occur at basal junction
of primary and secondary veins in section Macro-
catalpa (vs. at basal junction and along the midveins
LI: Catalpa phylogenetics

345
in section Catalpa). Trichomes are scale-like in
section Macrocatalpa (vs. global in section Catalpa).
Seeds are fimbriate all around in section Macro-
catalpa (vs. fimbriate terminally in section Catalpa)
(Elias & Newcombe, 1979). Britton (1918) elevated
section Macrocatalpa to the genus level based on leaf
morphology and geographic distribution. However,
because the two sections are each monophyletic and
are sister to each other, it is a matter of personal
preference whether or not to recognize Macrocatalpa
as a separate genus (Stevens, 1997). Here I adopt the
general view, recognizing them as sections Macro-
catalpa and Catalpa.
Within section Catalpa, the two North American
species have been suggested to be conspecific
(Warder, 1881; Roberts, 1902; Manning, 2000). They




Fig. 1. Strict consensus of 14 parsimonious trees of 441 steps inferred from ndhF sequences. Numbers above branches are bootstrap support
percentages. CI=0.82, RI=0.79. Numbers after species indicate accessions. EA, eastern Asia; NA, North America; WI, West Indies.
Journal of Systematics and Evolution Vol. 46 No. 3 2008 346


Fig. 2. nrDNA ITS trees. a, One of the 3 parsimonious trees of 117 steps (CI=0.88, RI=0.92). b, Single maximum likelihood tree (–ln=1513.9849).
Numbers at branches are bootstrap percentages. Asterisk indicates that the clade is absent in the strict consensus tree. EA, eastern Asia; NA, North
America; WI, West Indies.

share white flowers and other morphological charac-
ters (e.g., panicle inflorescence and greenish axillary
buds). Natural hybrids have been found in the south-
eastern Missouri where the two species have overlap-
ping distribution ranges (Brown, 1920). Nevertheless,
Catalpa speciosa flowers in the mid-June in Boston,
Massachusetts, while C. bignonioides blooms 10–15
days later (my personal observation). Catalpa speci-
osa has slightly fewer but larger flowers, fruits, and
seeds compared with C. bignonioides (Garman, 1912).
Catalpa speciosa and C. bignonioides are strongly
supported as a clade, reflecting their close affinity
(Figs. 1, 2).
Catalpa ovata, an eastern Asian species, is more
closely related to eastern North American species, C.
speciosa and C. bignonioides, than to other eastern
Asian species (Fig. 2). Interestingly, fertile hybrids
have been reported between C. ovata and either of the
two eastern North American species. The hybrids
grow faster and produce more seeds than either parent
(Sargent, 1889; Jones & Filley, 1920; Smith & Nich-
ols, 1941; Duffield & Snyder, 1958). All three species
have leaves with five basal primary veins and large
panicles (Paclt, 1952). Dode (1907) described Catalpa
duclouxii based on specimens collected in Yunnan,
China. This species differs from C. fargesii in the
glabrous leaf undersurface and inflorescence, and the
more branched inflorescence. Rehder (1913) treated
C. duclouxii as a variety of C. fargesii. Gilmour
(1936) further reduced C. duclouxii to a form of C.
fargesii. These treatments imply that C. duclouxii is
more closely related to C. fargesii than to C. bungei.
In the ITS and combined trees (Fig. 2), C. duclouxii
forms a clade that is sister to the clade containing C.
fargesii and C. bungei. Therefore, our results do not
support the close relationship of C. duclouxii and C.
fargesii, as suggested by Rehder (1913) and Gilmour
(1936).
The eastern Asian-eastern North American dis-
junction has attracted the attention of systematists and
biogeographers since the nineteenth century (Gray,
1846, 1859; Chaney, 1947; Axelrod, 1960; Wolfe,
1975; Boufford & Spongberg, 1983; Wu, 1983;
Tiffney, 1985a, b; Hong, 1993; Axelrod et al., 1998;
LI: Catalpa phylogenetics

347
Guo, 1999; Manchester, 1999; Qian & Ricklefs, 1999;
Wen, 1999; Donoghue et al., 2001; Stewart & Lister,
2001). Recent phylogenetic analyses of the disjunct
plant genera, albeit limited in number, have provided
valuable insights into pathways of migration and
patterns of diversification of plant lineages around the
Northern Hemisphere (Donoghue et al., 2001; Manos
& Donoghue, 2001; Xiang & Soltis, 2001; Donoghue
& Smith, 2004). However, in over 90 genera of the
disjunct distribution between eastern Asia and North
American disjunction, there are only six genera ex-
tending their distributions to the Caribbean islands. So
far, none of them have been studied in a phylogenetic
framework. In Catalpa, the WI species diverged from
the continental species, while North American species
may have come from Asia at a later time since they
are embedded in an Asian clade (Fig. 2). It remains to
be seen whether this is a general pattern in other
EAS-NA disjunct genera (e.g., Lyonia and Pieris).
4 Conclusions
Sequences of the chloroplast ndhF gene support
the sister relationship of Chilopsis and Catalpa, and
within Catalpa the sequence data recognize sections
Macrocatalpa and Catalpa. EAS species of section
Catalpa do not form a clade. Instead, Catalpa ovata is
more closely related to NA species than to other EAS
species, which form a clade. The WI species diverged
from EAS-NA continental species during the early
evolutionary history of Catalpa. Phylogenetic studies
of more disjunct genera with species in the WI are
warranted to test whether the early separation of the
WI species from continental lineages is a general
pattern.
Acknowledgements The author thanks Jeremy
LEDGER and Suzanne SHOUP for laboratory assis-
tance, Carl LEWIS of Fairchild Tropical Garden for
providing leaf material of Catalpa longissima and C.
macrocarpa, and Dr. Jun WEN for providing leaf
material of Catalpa duclouxii, and reading the early
version of the manuscript and offering constructive
suggestions.
References
Axelrod DI. 1960. The evolution of flowering plants. In: Tax S
ed. Evolution after Darwin. Chicago: The University of
Chicago Press. 1: 227–305.
Axelrod DI, Al-Shehbaz I, Raven PH. 1998. History of the
modern flora of China. In: Wu C-I, Wu S-K eds. Floristic
characteristics and diversity of East Asian plants:
Proceedings of the IFCD (1996). Beijing: Springer-Verlag
and China Higher Education Press. 43–55.
Baldwin BG, Sanderson MJ, Porter JM, Wojciechowski MF,
Campbell CS, Donoghue MJ. 1995. The ITS region of
nuclear ribosomal DNA: a valuable source of evidence on
angiosperm phylogeny. Annals of the Missouri Botanical
Garden 82: 247–277.
Baum DA, Small RL,Wendel JF. 1998. Biogeography and
floral evolution of baobabs (Adansonia, Bombacaceae) as
inferred from multiple data sets. Systematic Biology 47:
181–207.
Beillmann AP. 1946. The Catalpa—an interesting tree.
Missouri Botanical Garden Bulletin 34: 96–99.
Boufford DE, Spongberg SA. 1983. Eastern Asian-Eastern
North American phytogeographical relationships—a
history from the time of Linnaeus to the twentieth century.
Annals of the Missouri Botanical Garden 70: 423–439.
Britton NL. 1918. The relatives of Catalpa in the West Indies.
Journal of the New York Botanical Garden 19: 6–9.
Brown J. 1920. Catalpa speciosa in den Vereinigten Staaten
von Nordamerika. Mitteilungen der Deutschen
Dendrologischen Gesellschaft 29: 80–95.
Burk CJ, McMaster ND. 1988. The spread of Catalpa ovata G.
Don in western Massachusetts and its biogeographic
implications. Rhodora 90: 461–464.
Chaney RW. 1947. Tertiary centers and migration routes.
Ecological Monographs 17: 139–148.
Davis CC, Fritsch P, Li JH, Donoghue MJ. 2002. Phylogeny
and biogeography of Cercis (Fabaceae): evidence from
nuclear ribosomal ITS and chloroplast ndhF sequence
data. Systematic Botany 27: 289–302.
Dode LA. 1907. Notes dendrologiques. Bulletin da la Société
Dendrologigue de France 6: 190–209.
Donoghue MJ, Bell CD, Li JH. 2001. Phylogenetic patterns in
Northern Hemisphere plant geography. International
Journal of Plant Sciences 162 (6, Suppl.): S41–S52.
Donoghue MJ, Smith SA. 2004. Patterns in the assembly of
temperate forests around the Northern Hemisphere.
Philosophical Transactions of the Royal Society B:
Biological Sciences 359: 1633–1644.
Duffield JW, Snyder EB. 1958. Benefits from hybridizing
American forest tree species. Journal of Forestry 56:
809–815.
Duncan WH, Duncan MB. 1988. Trees of the southeastern
United States. Athens: The University of Georgia Press.
Elias TS, Newcombe LF. 1979. Foliar nectaries and glandular
trichomes in Catalpa (Bignoniaceae). Acta Botanica
Sinica 21: 215–224, plates I–IV.
Felsenstein J. 1985. Confidence limits on phylogenies: an
approach using the bootstrap. Evolution 39: 783–791.
Garman H. 1912. The catalpas and their allies. Kentucky
Agricultural Experiment Station Bulletin 164: 201–223.
Gentry A. 1980. Bignoniaceae—Part I. Flora Neotropica,
Monograph 25. New York: New York Botanical Garden.
Gentry A. 1992. Bignoniaceae—Part II. Flora Neotropica,
Monograph 25. New York: New York Botanical Garden.
Gentry A, Tomb AS. 1979. Taxonomic implications of
Bignoniaceae palynology. Annals of the Missouri
Botanical Garden 66: 756–777.
Gilmour JSL. 1936. Catalpa fargesii forma duclouxii. Curtis’s
Journal of Systematics and Evolution Vol. 46 No. 3 2008 348
Botanical Magazine 159: tab. 9458.
Gray A. 1846. Analogy between the flora of Japan and that of
the United States. American Journal of Science and Arts
(ser. 2) 2: 135–136.
Gray A. 1859. Diagnostic characters of new species of
phanerogamous plants, collected in Japan by Charles
Wright, botanist of the U. S. North Pacific Exploring
Expedition, with observations upon the relations of the
Japanese flora to that of North America, and of other parts
of the northern temperate zone. Memoirs of the American
Academy of Arts and Sciences (N. S.) 6: 377–453.
Guo QF. 1999. Ecological comparisons between eastern Asia
and North America: Historical and geographical
perspectives. Journal of Biogeography 26: 199–206.
Henrickson J. 1985. A taxonomic revision of Chilopsis
(Bignoniaceae). ALISO 11: 179–197.
Hong DY. 1993. Eastern Asian-North American disjunctions
and their biological significance. Cathaya 5: 1–39.
Jones DF, Filley WO. 1920. Teas’s hybrid Catalpa. Journal of
Heredity 11: 16–24.
Li HL. 1952. Floristic relationships between eastern Asia and
eastern North America. Transactions of the American
Philosophical Society 42: 371–429.
Li J, Shoup S, Elias TS. 2006. Molecular confirmation of
intergeneric hybrid ×Chitalpa tashkentensis (Bignoni-
aceae). HortScience 41: 1–3.
Little EL. 1979. Checklist of United States trees (native and
naturalized). Washington DC: Forest Service, USDA.
Manchester SR. 1999. Biogeographical relationships of North
American Tertiary floras. Annals of the Missouri Botanical
Garden 86: 472–522.
Manning SD. 2000. The genera of Bignoniaceae in the south-
eastern United States. Harvard Papers in Botany 5: 1–79.
Manos PS, Donoghue MJ. 2001. Progress in Northern
Hemisphere biogeography: An introduction. International
Journal of Plant Sciences 162: S1–S2.
Olmstead RG, Kim K-J, Jansen RK, Wagstaff SJ. 2000. The
phylogeny of the Asteridae sensu lato based on chloroplast
ndhF gene sequences. Molecular Phylogenetics and
Evolution16: 96–112.
Olmstead RG, Sweere JA. 1994. Combining data in
phylogenetic systematics: an empirical approach using
three molecular data sets in the Solanaceae. Systematic
Biology 43: 467–481.
Paclt J. 1952. Synopsis of the genus Catalpa (Bignoniaceae)
III. Candollea 13: 241–285.
Posada D, Crandall KA. 1998. MODELTEST: testing the
model of DNA substitution. Bioinformatics (Applications
Note) 14: 817–818.
Qian H, Ricklefs RE. 1999. A comparison the taxonomic
richness of vascular plants in China and the United States.
The American Naturalist 154: 160–181.
Rehder A. 1913. Bignoniaceae. In: Sargent CS ed. Plantae
Wilsonianae I. Cambridge: Harvard University Press.
Roberts HF. 1902. The hardy Catalpa. Kansas Agricultural
Experiment Station Bulletin 108: 99–214.
Rusanov NF. 1964. On the intergeneric hybrids of Catalpa and
Chilopsis. Byulleten’ Glavnogo Botaniceskogo Sada. 55:
44–47. (in Russian)
Sargent CS. 1889. A hybrid Catalpa. Garden & Forest 2:
303–305.
Smith EC, Nichols CJ. 1941. Species hybrids in forest trees.
Journal of the Arnold Arboretum 22: 443–454.
Soltis DE, Soltis PS, Doyle JJ eds. 1998. Molecular systematics
of plants II: DNA sequencing. Norwell, MA: Kluwer
Academic Publishers.
Spangler RE, Olmstead RG. 1999. Phylogenetic analysis of
Bignoniaceae based on the cpDNA gene sequences rbcL
and ndhF. Annals of the Missouri Botanical Garden 86:
33–46.
Stevens PF. 1997. What kind of classification should the
practising taxonomist use to be saved? In: Dransfield J,
Coode MJE, Simpson DA eds. Plant diversity in Malesia
III. Kew: Royal Botanical Gardens.
Stewart JR, Lister AM. 2001. Cryptic northern refugia and the
origins of the modern biota. Trends in Ecology &
Evolution 16: 608–613.
Swofford DL. 2002. PAUP*. Phylogenetic analysis using
parsimony (* and other methods). Sunderland, MA:
Sinauer Associates.
Thomas RD, Allen CM. 1996. Atlas of the vascular flora of
Louisiana. Vol. 2. Dicotyledons: Acanthaceae-
Euphorbiaceae. Baton Rouge, Louisiana: Louisiana
Department of Wildlife and Fisheries.
Tiffney BH. 1985a. Perspectives on the origin of the floristic
similarity between eastern Asia and eastern North
America. Journal of the Arnold Arboretum 66:73–94.
Tiffney BH. 1985b. The Eocene North Atlantic land bridge: its
importance in tertiary and modern phytogeography of the
northern hemisphere. Journal of the Arnold Arboretum 66:
243–273.
Wang W-C. 1990. Catalpa Scop. In: Flora Reipublicae
Popularis Sinicae. Beijing: Science Press. 69: 13–18.
Warder JA. 1881. The western Catalpa, a memoir of the
shavanon, or the Catalpa speciosa (Engelmann). Journal
of the American Agricultural Association 1: 79–102.
Wen J. 1999. Evolution of eastern Asian and eastern North
American disjunct distributions in flowering plants.
Annual Review of Ecology and Systematics 30: 421–455.
Weniger D. 1996. Catalpa (Catalpa bignonioides,
Bignoniaceae) and bios d’arc (Maclura pomifera,
Moraceae) in early Texas records. Sida 17: 231–242.
White TJ, Bruns T, Lee S, Taylor J. 1990. Amplification and
direct sequencing of fungal ribosomal RNA genes for
phylogenetics. In: Innis M, Gelfand D, Sninsky J, White T
eds. PCR Protocols: A guide to methods and applications.
San Diego: Academic Press. 315–322.
Wolfe JA. 1975. Some aspects of plant geography of the
northern hemisphere during the Late Cretaceous and
Tertiary. Annals of the Missouri Botanical Garden 62:
264–279.
Wu CI. 1983. On the significance of Pacific intercontinental
discontinuity. Annals of the Missouri Botanical Garden
70: 577–590.
Xiang QY, Soltis DE. 2001. Dispersal-vicariance analysis of
intercontinental disjuncts: historical biogeographical
implications for angiosperms in the Northern Hemisphere.
International Journal of Plant Sciences 162: S29–S39.