The Eastern Asian and Eastern and Western North American
Floristic Disjunction: Congruent Phylogenetic Patterns
in Seven Diverse Genera
Qiu-Yun Xiang,1DouglasE. Soltis, and Pamela S. Soltis
Department of Botany, Washington State University, Pullman, Washington 99164–4238
Received August 18, 1997; revised January 5, 1998
One of the most remarkable examples of interconti-
nental disjunction of the North Temperate F lora in-
volves eastern Asia and eastern and western North
America. Although there has been considerable inter-
est in this phytogeographic pattern for over 150 years
(e.g., Gray, 1859; L i, 1952; Graham, 1972; Boufford and
Spongberg, 1983; Wu, 1983; Tiffney, 1985a, 1985b), rela-
tionships among taxa displaying the disjunction re-
main obscure. Understanding phylogenetic relation-
ships is, however, aprerequisite
biogeographic analyses of this distributional pattern.
To understand better the relationships of taxa display-
ing this intercontinental disjunction, phylogenetic
analyses were conducted using a variety of DNA data
sets for species of four genera (Cornus, Boykinia,
Tiarella, and Trautvetteria) that occur in eastern Asia,
eastern North America, and western North America.
An area cladogram was constructed for each of the
four genera, all of which show a similar pattern of
relationship: the eastern Asian species are sister to all
North American species. An identical phylogenetic
pattern is also found in three other taxa exhibiting this
disjunction (Aralia sect. Aralia, Calycanthus, and Adi-
antum pedatum). T he congruent phylogenetic pattern
found in these seven diverse genera raises the possibil-
ity of a common origin of the eastern Asia, eastern and
western North America disjunction. T he data are in
agreement with the long-standing hypothesis that this
well-known floristic disjunction represents the frag-
mentation of a once continuous Mixed Mesophytic
forest community and suggest that the disjunction
may have involved only two major vicariance events:
an initial split between E urasia and North America,
followed by the isolation of floras between eastern and
western North America. However, congruence be-
tween phylogenies and geographic distributions does
not necessarily indicate an identical phytogeographic
history. Taxa exhibiting the same phylogenetic pattern
may have originated at different geological times.
Analysis of divergence times using the molecular clock
indicates that species of Cornus, Boykinia, and Calycan-
thus may have diverged at different geological times,
suggesting that the floristic disjunction involving eastern
Asia and North America may not be simple; it may have
involved multiple historical events at very different
geological times in different genera.
?1998 Academic Press
The flora of the North Temperate Zone exhibits
striking intercontinental floristicsimilarities.Thissimi-
larity is most remarkable between eastern Asia and
eastern North America, and toa lesser extent in twoor
more of the following five areas: eastern Asia, western
Asia, eastern North America, western North America,
and southeastern Europe. Floristic elements display-
ing intercontinental disjunction in the North Temper-
ateZonehavegenerally been consideredTertiary relicts
that prior tothelateMioceneweremorewidely distrib-
uted across Laurasia, forming part of a Mixed Meso-
phytic forest community (Li, 1952; Wolfe and Leopold,
1967; Wolfe, 1969, 1972, 1975; Wood, 1971, 1972;
Graham, 1972). Following this hypothesis, the disjunc-
tion arose following geographic and climatic changes,
including the separation of North America and Eurasia
duetocontinental drift, theclosing of theTurgai Straits
in the Old World during the Oligocene, the Tertiary
orogenies of western North America, and worldwide
climaticfluctuations through theTertiary. Theseevents
eliminated plants from many areas, particularly Eu-
rope and central North America (Graham, 1972; Leo-
pold and MacGinitie, 1972; Tiffney, 1985a).
The disjunct distributional patterns of the North
Temperate Flora, particularly those involving eastern
Asia and NorthAmerica, havebeen thesubject of study
for more than 150 years (e.g., Gray, 1859, 1878; Hu,
1935; Chaney, 1947; Li, 1952, 1972; Koyama and
Kawano, 1964; Graham, 1972; Hara, 1972; Wolfe, 1972,
1To whom correspondence and reprint requests should be ad-
dressedat IdahoStateUniversity, Box 8007 Pocatello, ID 83209. Fax:
(208) 236–4570. E-mail: firstname.lastname@example.org. tree.
MOLECULAR PHYLOGENETICS AND EVOLUTION
Vol. 10, No. 2, October, pp. 178–190, 1998
ARTICLE NO. MPE980524
Copyright?1998 by Academic Press
All rights of reproduction in any form reserved.
1975, 1981, 1985; Boufford and Spongberg, 1983; Iltis,
1983; Hong, 1983, 1993; Hsu, 1983; Koyama, 1983;
Tamura, 1983; Wu, 1983; Tiffney, 1985a, 1985b; Bouf-
ford, 1992). Despite long-standing interest, phyloge-
netic relationships among taxa displaying the disjunc-
tion have remained obscure, and the origin of the
disjunction has been controversial. Major disagree-
ments exist regarding: (1) thepattern of relationship (if
any) among plants exhibiting the disjunction, (2)
whether the disjunction had a single origin or involved
multiple biogeographic events at different times in
different taxa, and(3) theproposedgeographicorigin(s)
of the disjunct taxa. Before these and other issues
regarding this prominent biogeographic disjunction
can be properly evaluated, a clear understanding of the
phylogeny of taxa exhibiting the disjunct patterns is
Wefocused on taxa showing a disjunct distribution in
eastern Asia and both eastern and western North
America.Approximately 30 genera have closely related
species occurring in these three areas (Wood, 1971,
1972; Li, 1972; Wu, 1983). There is no consensus of
opinion as to how plants found in all three areas are
related to one another; authors have suggested differ-
ent patterns for different taxa. In Trautvetteria, Tamura
(1983) suggested that the closest relationship is be-
tween the eastern Asian and western North American
species. In contrast, in Amsonia, Gaultheria, Osmo-
rhiza, and Styrax, Wood (1972) considered the closest
relationship to be between the eastern Asian and
eastern North American taxa. These hypotheses were
based solely on morphological similarities, and no
phylogenetic analyses wereconducted.
Only a few genera occurring in eastern Asia and both
eastern and western North America have been studied
phylogenetically, including Mitella (Soltis et al., 1991;
Soltis and Kuzoff, 1995), Hydrangea (Soltis et al.,
1995), Adiantum pedatum L. (Paris, 1991; Paris and
Haufler, 1994), Trillium (Kato et al., 1995), Calycan-
thus (Wen et al., 1996), and Aralia sect. Aralia (Aralia-
ceae) (Wen et al., 1996). Phylogenetic studies have
revealed that Hydrangea and Mitella, genera long
considered ‘‘classic’’ examples of this floristic disjunc-
tion, are polyphyletic. Hence these genera are not
useful models for the study of this disjunction. These
findings clearly demonstrate the importance of a sound
phylogenetic framework in inferring biogeographic his-
tory; comparison of taxa that are not truly closest
relatives may lead to erroneous biogeographic conclu-
sions. In contrast, phylogenetic analysis of Trillium
indicated that, although monophyletic, phytogeo-
graphic hypotheses will be best considered within
discrete subclades of the genus (Kato et al., 1995), a
task that requires greater taxon density than presently
available. Analyses of Calycanthus (three species with
one in each area) and Aralia sect. Aralia (nine species,
one in eastern North America, one in western North
America, and seven in eastern Asia) suggest that
species from eastern and western North America are
sisters, which arein turn thesister of theAsian species.
Clearly, more taxa showing this disjunct distributional
pattern need to be examined phylogenetically to im-
prove our understanding of the eastern Asia, eastern
and western North America disjunction. We therefore
conducted molecular phylogenetic analyses of four gen-
era, the big-bracted dogwoods of Cornus (Cornaceae),
Boykinia, and Tiarella (Saxifragaceae), and Trautvette-
ria (Ranunculaceae), groups that currently occur only
in eastern Asia, eastern North America, and western
North America. Our goals were to: (1) estimate phylog-
enies for these four genera, (2) determine whether
there are general phylogenetic patterns for diverse
genera showing this distributional pattern; (3) con-
struct area cladograms based on the molecular phylog-
enies of these genera to elucidate relationships among
eastern Asia, eastern North America, and western
North America; (4) gain initial insight into the geo-
graphic origin of taxa exhibiting this disjunction to
improve our understanding of the North Temperate
MATERIALS AND METHODS
Molecular Phylogenetic Analyses
Several DNA regions weresequenced toinfer phylog-
eny, although not all regions were analyzed for each
genus: the chloroplast genes rbcL and matK and the
internal transcribed spacers of the nuclear ribosomal
RNA genes (ITS regions). Methods of amplification and
sequencing followed Morgan and Soltis (1993) and
Xiang et al. (1993) for rbcL, J ohnson and Soltis (1994,
1995) for matK, and Baldwin (1992) and Soltis and
Kuzoff (1995) for ITS. Previously published cpDNA
restriction site variation was also used for phylogeny
estimation (Soltis et al., 1991, 1993; Xiang et al., 1996).
The general methods of phylogenetic analysis were
as follows unless specified in the figure legends. For
each genus, several outgroups were used. In addition,
both broad phylogenetic analyses involving related
genera and more focused analyses involving only the
study genera were conducted. Outgroups for the nar-
row analyses were chosen based on results from broad
analyses that identified sister taxa of the study group.
The pattern of phylogenetic relationships within each
genus was thesamein all analyses.
Data were analyzed with PAUP 3.1.1 (Swofford,
1993) using Fitch parsimony. For Cornus and Boykinia,
a branch-and-bound search was conducted, and for
Tiarella and Trautvetteria, the exhaustive search op-
tion was used. To evaluate the relative support of
relationships revealed in the most parsimonious trees,
bootstrap analysis (Felsenstein, 1985) with 100 repli-
cates was performed, and Bremer support (or decay
index) (Bremer, 1988) was estimated following Eer-
PHYLOGENETIC PATTERNS OF DISJ UNCT TAXA
nisseand Kluge(1993).Additional background data for
each genus analyzed arepresented below.
Cornus.Cornus(thedogwoods) isa genusofapproxi-
mately 55 species, within which the big-bracted dog-
woods form a monophyletic group (Xiang et al., 1993,
1996; Xiang and Soltis, in press). This group consists of
seven closely related but morphologically variable spe-
cies, with C. florida L. distributed in eastern North
America, C. disciflora Moc. & Sesse ´in Central America,
C. nuttallii Audubon in western North America, and
C. capitata Wall., C. kousa Hance, C. hongkongensis
Hemsley, and C. multinervosa (Pojarkova) Q. Y. Xiang
in eastern Asia (Xiang, 1987). Themorphological varia-
tion among species of the big-bracted dogwoods paral-
lels their geographic distributions. All eastern Asian
species are morphologically very similar and produce
compound fruits in heads. These species also form a
distinct clade in a recent phylogenetic analysis using
morphological characters (Murrell, 1993).All American
species produce separate fruits in clusters. Cornus
florida from eastern North America and the eastern
Asian species have four large, showy bracts, whereas
C. nuttallii from western North America has six large
showy bracts; C. disciflora from Central America has
four bracts that absciseearly without having expanded.
Phylogenetic analyses employed cpDNA restriction
sites and a combined data set of rbcL–matK sequences–
cpDNA restriction sites (1440 bp of rbcL, 1212 bp of
matK, and 242 restriction sites). All three American
species, C. florida, C. disciflora, and C. nuttallii, and
two representatives of the eastern Asian clade,
C. capitata and C. kousa, were included in the analysis
of cpDNA restriction sites. All of these species except
C. disciflora (which was not included because rbcL and
matK could not be amplified due to the degradation of
DNA subsequent to the earlier restriction site study)
were then included in a combined analysis of rbcL–
matK sequences and cpDNA restriction sites. All taxa
included in this analysis of combined data sets have at
least two of the three data sets available. On the basis
of the results of previous studies of Cornus and its
closest relatives (Xiang et al., 1996, 1998a), two dwarf
dogwoods (C. canadensis and C. unalaschkensis) and
three cornelian cherries (C. mas, C. officinalis, and C.
sessilis) were selected as outgroups. Molecular data
were from our recent studies (Xiang et al., 1993, for
rbcL sequences; 1996, for restriction sites; and 1998a,
for matK sequences), except for the rbcL sequence of
Cornus nuttallii, which was generated in this study
Boykinia. Boykinia contains seven species, with
B. aconitifolia Nutt. the only species from eastern North
America, B. lycoctonifolia (Maxim.) Engl. the single
species in eastern Asia, and five species, B. intermedia
(Piper) G. N. J ones, B. major A. Gray, B. occidentalis
Torrey & Gray, B. rotundifolia Parry, and B. richardso-
nii (Hook.) Rothrock, in western North America (Gor-
nall and Bohm, 1985; Soltis et al., 1993). The mono-
phyly of Boykinia has been demonstrated in several
previous phylogenetic analyses of Saxifragaceae s. s.
(Soltis et al., 1993, 1996; J ohnson and Soltis, 1995). We
focused on relationships within the genus and con-
ducted phylogenetic analyses using cpDNA restriction
sites and nuclear ITS sequences separately. All species
of the genus except B. richardsonii, a high polyploid for
which suitable DNA was not available (see Soltis et al.,
1993), were included. Chloroplast DNA restriction site
data were from Soltis et al. (1993), and ITS sequences
were from Soltis et al. (1996). Suksdorfia and Bolandra
wereused as outgroups based on theresults of previous
analyses of Saxifragaceae s. s. (Soltis et al., 1993).
Because the cpDNA and ITS trees revealed different
TABL E 1
Sources of Material and Molecular Data Generated in T his Study for Phylogenetic Analyses
TaxaData Voucher or sources Genebank Accession Numbers
C. nuttallii Audubon
T. polyphylla D. Don
T. carolinensis (Walt.) Vail
rbcL Arn. Arb. No. 573-73-A (WS)AF006833
ITS-1, ITS-2 Soltis 2555 (WS) AF006834, AF015444
Hoot 92018, F, Hoot Garden
Hoot 92018, F, Hoot Garden
Ohba A3018, NikkoBot. Gard., J apan
Ohba A3018, NikkoBot. Gard., J apan
T. grandis Nutt.
T. japonica Sieb. & Zucc.
M. minimus L.ITS-1, ITS-2
Chase532, K, RBG, Kew
Chase532, K, RBG, Kew
Note. Abbreviations of UCBG, Arn. Arb., and RBG represent University of California Botanical Garden, the Arnold Arboretum, and the
Royal Botanical Gardens, Kew, respectively. Xiang and Soltis vouchers aredeposited at WS.
XIANG, SOLTIS, AND SOLTIS
relationships among the North American taxa, the two
data sets werenot combined intoa singleanalysis.
Tiarella.Tiarella has three species: T. cordifolia L.
in eastern North America, T. trifoliata L. with three
varieties (var. trifoliata, var. unifoliata (Hook.) Kurtz.,
and var. laciniata (Hook.) Wheel.) in western North
America, and T. polyphylla D. Don in eastern Asia.
Previous analyses of several chloroplast markers sug-
gested that Tiarella is polyphyletic (Soltis et al., 1991;
Xiang, 1995). These analyses indicate that the eastern
North American species (T. cordifolia) is more closely
related to Heuchera than it is to the other species of
Tiarella. Morphology and nuclear ITS sequence data
suggest, in contrast, that Tiarella is monophyletic
(Soltis and Kuzoff, 1995). The totality of data indicates
that T. cordifolia has apparently captured the chloro-
plast of a species of Heuchera and that the cpDNA-
based topologies do not reflect organismal relation-
ships. As reviewed elsewhere (Soltis and Kuzoff, 1995),
this chloroplast capture scenario is further suggested
by the presence of naturally occurring intergeneric
hybrids between T. cordifolia and species of Heuchera.
Because the chloroplast genome is providing an
inaccurate assessment of organismal relationships in
Tiarella, our phylogenetic hypothesis is based solely on
nuclear sequences. In this study, weconductedphyloge-
netic analyses of the three species of Tiarella using ITS
sequences. A broad analysis of ITS sequences for the
entireHeuchera group (a cladecomprising Tiarella and
eight other genera of Saxifragaceae; Soltis et al., 1991,
1993) was first performed using Darmera and Rodger-
sia as outgroups. Sequences of ITS werefromSoltis and
Kuzoff (1995), except that for T. polyphylla, which was
obtained in this study. Because our broad analysis of
ITS sequences indicates that Tiarella is monophyletic
(tree not shown), a second narrow analysis of ITS
sequences was conducted for Tiarella using an exhaus-
tive search. Conimitella williamsii, Mitella diversifo-
lia, and Mitella stauropetala were used as outgroups
for this focused analysis because of their close phyloge-
netic relationship toTiarella.
Trautvetteria.Trautvetteria consists of three spe-
cies: T. carolinensis (Walt.) Vail from eastern North
America, T. grandisNutt. fromwestern NorthAmerica,
and T. japonica Sieb. & Zucc. from eastern Asia.
Sequences of matK and ITS were obtained for these
three species, as well as for the outgroup Myosurus
minimus L. Myosurus was chosen as the outgroup
based on the results of broad phylogenetic analyses of
Ranunculaceae (Hoot, 1995, personal communication)
that indicate that Myosurus is a close relative of
Trautvetteria. Of the sequencing primers designed by
J ohnson and Soltis (1994, 1995) for matK, only two,
1168R and 1470R, could be used in this study. These
primers provided approximately 600 base pairs (bp) of
matK sequence for the species studied. Sequences of
both theITS-1 and ITS-2 regions werealsoobtained for
Trautvetteria andMyosurus. Separateanalyses ofmatK
and ITS sequences were performed. Because ITS and
matK sequences yielded identical topologies, we com-
bined these data sets and conducted subsequent analy-
ses on thecombined data matrix.
Area cladograms were constructed for the big-
bracted dogwoods of Cornus, Boykinia, Tiarella, and
Trautvetteria by replacing terminal taxon names on a
cladogram with the distributional area of the respec-
tive taxon (see Platnick and Nelson, 1978; Nelson and
Platnick, 1981; Humphries and Parenti, 1986). The
area cladograms for the four genera were compared,
and a congruent general area cladogram was constructed
North America, and western North America.
Center of Origin
Establishing the center of origin for a group without
phylogenetic information (e.g., using the ‘‘age-and-
area’’ hypothesis; see Brown and Gibson, 1983; Fu-
tuyma, 1986) is often problematic. We attempted to do
this using a phylogenetic approach using MacClade
3.05 (Maddison andMaddison, 1992) tomapthecharac-
ter ‘‘area’’ onto the topology in the most parsimonious
manner toinfer thegeographicarea of theancestor (the
root node) of each study group. In many cases, however,
conducting this mapping exercise was so straightfor-
ward that it could bedoneby hand.
Topologies resulting from broad phylogenetic analy-
ses that clearly depict the sister group of the study
group (thesetrees arenot shown here) wereused as the
basis for determining the outgroups used in the topol-
ogy for ‘‘area’’ mapping. For Cornus, the topologies
resulting from broad analyses of cpDNA restriction
sites for the entire genus and a combined rbcL–matK
data set representing the entire Cornaceae were used
for determining the sister group of the big-bracted
dogwoods (Xiang et al., 1993, 1998a). For Boykinia,
both the ITS and cpDNA restriction site topologies for
the entire ‘‘Boykinia group’’ (a clade of six genera)
(Soltis et al., 1993, 1996), as well as cpDNA restriction
site and matK topologies representing the entire Saxi-
fragaceae s. s. (Soltis et al., 1993, 1996), were used. For
Tiarella, we used the ITS tree for the entire Heuchera
group (a clade of nine genera) (Soltis and Kuzoff, 1995).
For Trautvetteria, we used the combined atpB–rbcL–
18S rDNA tree for Ranunculaceae (Hoot, 1995). The
geographic distribution of the outgroup was assigned
based on the known distribution of all members of the
outgroup (not just the several outgroup species de-
picted in the figures). Therefore, geographic areas of
species shown in thephylogenetictrees donot necessar-
ily represent the geographic areas of the entire out-
group genera. For example, the sister group of the
PHYLOGENETIC PATTERNS OF DISJ UNCT TAXA
big-bracted dogwoods was indentified as the dwarf
dogwoods by the broad analysis for the genus. The
dwarfdogwoods havefour species distributedcircumbo-
really and extending to Burma (see Xiang et al., 1996).
In the phylogenetic trees (Fig. 1), only two North
American species, C. canadensis and C. unalaschken-
sis, are shown. In the geographic ‘‘area mapping’’
analysis, the geographic area of the outgroup of the
big-bracted dogwoods was considered as circumboreal,
instead of North America.
We also performed this same exercise for Calycan-
thus and Aralia sect. Aralia, taxa investigated phyloge-
netically by others (Wen et al., 1996). For Calycanthus,
the broad phylogenetic analysis of rbcL sequences
representing magnoliids (Qiu et al., 1993) was em-
ployed. For Aralia sect. Aralia, the cpDNA restriction
site topology of the section, using Aralia sect. Dimorphan-
thusasthesister group(Wen etal., 1996), wasused.
Cornus.Phylogenetic analysis of cpDNA restriction
sites for thebig-bracted dogwoods reveals relationships
(Fig. 1a) identical to those found in broad analyses of
the entire genus Cornus (Xiang et al., 1996). That is,
theeasternAsian speciessampled(C.capitata, C.kousa)
form a clade and are the sister group of a clade
containing all American species. Analysis of the com-
bined rbcL–matK–cpDNA restriction site data set for
the big-bracted dogwoods reveals the same general
topology observed with restriction sites alone(although
now C. disciflora is omitted; see above) (Fig. 1b). The
eastern Asian species (C. capitata, C. kousa) form a
clade and are sister to the North American species
analyzed, C. florida and C. nuttallii.
Boykinia. The phylogenetic analyses of cpDNA re-
striction sites and ITS sequences for Boykinia both
place the eastern Asian species, B. lycoctonifolia, as
sister to a clade that comprises all of the North
American species (Figs. 2a and 2b). Within this North
American clade, the ITS tree places B. rotundifolia
from western North America as the sister to a clade
containing B. aconitifolia from eastern North America
and the remaining western North American species
(Fig. 2b), whereas the cpDNA tree shows a trichotomy
among B. aconitifolia, B. rotundifolia, and the remain-
ing species (Fig. 2a).
Tiarella and Trautvetteria.
sis of ITS sequences for Tiarella and analyses of ITS
and matK sequences for Trautvetteria indicate that, for
both genera, the two North American species are more
closely related to each other than either is to the
eastern Asian species (Figs. 3 and 4).
The four area cladograms based on the molecular
phylogenies of the big-bracted species of Cornus,
Tiarella, Boykinia, and Trautvetteria are congruent in
showing that eastern and western North America are
more closely related to each other than either is to
eastern Asia (Fig. 5).
Center of Origin
Our attempts to infer the centers of origin of these
four genera by mapping geographic area onto topolo-
F IG. 1.
of the cpDNA restriction site data using cornelian cherries (C. mas, C. officinalis, and C. sessilis) and the dwarf dogwoods (C. canadensis and
C. unalaschkensisas) as outgroups (length ? 175 steps; CI ? 0.961, excluding uninformativecharacters, RI ? 0.982). Ingroup is designated in
bold. The number of restriction site mutations supporting each clade is given above each branch, along with the decay value (in parentheses);
bootstrap values aregiven below branches. Thedashed linerepresents thebranch not recognized in all thethreeshortest trees. (b) Thesingle
shortest tree resulting from the combined analysis of rbcL-matK sequences and restriction sites of cpDNA for the big-bracted dogwoods
(length ? 405 steps; CI ? 0.938, excluding uninformativecharacters, RI ? 0.962). Ingrouparedesignatedin bold. Numbers on branches areas
designated for (a).
Phylogenetic analyses of the big-bracted dogwoods of Cornus. (a) One of the three shortest trees resulting from exhaustive search
XIANG, SOLTIS, AND SOLTIS
gies have produced several outcomes that are largely
correlated with thegeographic distribution of thesister
group of the study group (see Figs. 1–4). For example,
all broad phylogenetic analyses of the Boykinia group,
as well as of Saxifragaceaes. s. (Soltis et al., 1993, 1995,
1996), indicate that the sister taxa of Boykinia are
Bolandra and Suksdorfia, genera confined to western
North America. Hence, for Boykinia, the area-mapping
suggests that the center of origin is western North
America. In Cornus, in contrast, the immediate sister
group of the big-bracted dogwoods is the dwarf dog-
woods, taxa with a circumboreal distribution. Thus, the
place of origin for the big-bracted dogwoods is equivo-
cal. Similarly, for Trautvetteria, the sister genera, Myo-
surus and Ranunculus (Hoot, 1995), are both widely
distributed in the Northern Hemisphere. Thus, once
again thegeographicorigin for thegenus is indicatedas
uncertain. For Tiarella, no well-supported sister group
has been identified (Soltis and Kuzoff, 1995). Finding
the sister group of Tiarella is complicated because
extensive chloroplast capture in the clade of genera to
which it belongs has rendered cpDNA data useless for
phylogenetic inference, and ITS sequence data provide
insufficient resolution of the relationships of Tiarella.
Thus, the center of origin for the genus also remains
For Calycanthus, the sister group is Chimonanthus
(the only other member of Calycanthaceae), a genus
restricted to eastern Asia (Qiu et al., 1993). Thus, Asia
is implicated as the center of origin for Calycanthus.
The sister group of Aralia sect. Aralia is Aralia sect.
Dimorphanthus (Wen et al., 1996), which occurs in both
eastern Asia and eastern North America. Although the
site of origin of section Aralia is equivocal based on the
F IG. 2.
Boykinia (length ? 57 steps; CI ? 0.911, excluding uninformative characters, RI ? 0.965). Ingroup are designated in bold. Numbers on
branches are as designated for Fig. 1a. (b) One of the two shortest trees resulting from phylogenetic analysis of ITS sequences for Boykinia
(length ? 236 steps; CI ? 0.747, excluding uninformativecharacters, RI ? 0.802). Ingrouparedesignatedin bold. Numbers on branches areas
designated for Fig. 1a. Dashed lines represent branches not recognized in both shortest trees; all other branches are present in both most
Phylogenetic analyses of Boykinia. (a) The single shortest tree resulting from phylogenetic analysis of cpDNA restriction sites for
F IG. 3.
analysis of ITS sequences of Tiarella (length ? 39 steps, CI ? 0.952,
excluding uninformative characters, RI ? 0.971). Ingroup are desig-
nated in bold. Numbers on branches areas designated for Fig. 1a.
The single shortest tree resulting from phylogenetic
F IG. 4.
analyses of ITS, matK, and combined ITS–matK sequences, respec-
tively, for Trautvetteria. The single ITS shortest tree has a length of
120 steps, a CI of 0.974, excluding uninformativecharacters, anda RI
of 0.971; thesinglematK shortest treehas a length of 46 steps, a CI of
1.000, excluding uninformative characters, and a RI of 1.000; the
single ITS–matK tree has a length of 166 steps, a CI of 0.983,
excluding uninformataive characters, and a RI of 0.982. Ingroup are
designated in bold. Numbers on branches are as designated for Fig.
1a; thosein italic arefor thematK treeand thosein roman arefor the
The identical shortest trees resulting from phylogenetic
PHYLOGENETIC PATTERNS OF DISJ UNCT TAXA
distribution of its sister group, an origin in Asia is
suggested by the distributions of the basal lineages
within section Aralia. The first-branches of section
Aralia terminatein species from eastern Asia.
Pattern of relationships.
four genera (the big-bracted species of Cornus, Boyki-
nia, Tiarella, and Trautvetteria) occurring disjunctly in
eastern Asia, eastern North America, and western
North America reveal a single pattern of biogeographic
relationships among species (Figs. 1–4). In all four
genera, species from eastern North America are the
sister group to species from western North America,
and the Asian species are the sister group to all North
American species. In contrast, intuitiveinferences (i.e.,
nonphylogenetic) based on other sources of data sug-
gested alternative relationships or did not resolve
relationships within each genus. For example, the
three species of Tiarella have very similar karyotypes,
and each has a unique set of flavonoid constituents
(Soltis and Bohm, 1984). These characters did not
provide any synapomorphies within Tiarella and sug-
gested that all three species are equally distinct. Non-
cladistic inferences of relationships in Trautvetteria
based on morphological characters suggested that the
eastern Asian species (T. japonica) and the western
North American species (T.grandis) are more closely
related toeach other than either is totheeastern North
American species (T. carolinensis) (Tamura, 1983).
This conclusion conflicts with the phylogenetic analy-
ses presented here.
The pattern of phylogenetic relationships reported
Phylogenetic analyses for
here for four genera is similar to that for Aralia sect.
Aralia (Araliaceae) (Wen et al., 1996), Calycanthus
(Calycanthaceae) (Wen et al., 1996), and Adiantum
pedatum L. (Adianthaceae) (Paris, 1991; Paris and
Haufler, 1994), all with species or populations occur-
ring in eastern Asia, eastern North America, and
western NorthAmerica. Therefore, phylogenetic analy-
ses indicate that the big-bracted species of Cornus,
Boykinia, Tiarella, Trautvetteria, Aralia, Calycanthus,
and Adiantum pedatum, a diverse array of taxa, all
show a similar pattern of relationships among species:
theeastern andwestern NorthAmerican species consis-
tently are sisters, with the Asian representative(s) in
turn sister totheNorth American clade.
It is not clear how general the phylogenetic pattern
observed here may be among the approximately 30
genera that exhibit this disjunction.Additional phyloge-
netic analyses of taxa exhibiting this disjunction are
therefore encouraged. It is alsonoteworthy that a large
number of genera exhibit a slightly more complex
disjunction pattern than that discussed here. That is,
in addition to having representatives in eastern Asia
and both eastern and western North America, some
genera also have member taxa in Europe (e.g., Aescu-
lus, Chrysosplenium). Nonetheless, several recent phy-
logenetic analyses indicate that a phylogenetic pattern
similar to that noted here may exist for both Aesculus
L. (Hippocastanaceae; ITS and matK sequence data;
Xiang et al., in 1998b) and Chrysosplenium (Hibsch-
J etter et al., unpublished) in that the North American
species form a cladethat is sister totheOld World taxa.
Congruent biogeographic pattern versus shared phy-
togeographic history.Congruence between phyloge-
netic topologies and geographic distribution of lineages
has been used as a basis for inference of shared
historical biogeography (Platnick and Nelson, 1978;
Nelson and Platnick, 1981; Humphries and Parenti,
1986; Sober, 1988; Oosterbroek and Arntzen, 1992).
Based on this principle, the consistent pattern of
phylogenetic relationship observed among the seven
diverse taxa compared here may indicate a common
biogeographic history. These data can, for example, be
viewed as support for the long-standing hypothesis
that the disjunction in eastern Asia, eastern North
America, and western North America represents the
fragmentation of a once-continuous plant community
(e.g., Gray, 1878; Hu, 1935; Chaney, 1947, 1959; Li,
1952; Wood, 1971, 1972). For example, Gray (1878)
proposedthat a continuous flora existedacross thehigh
latitudes of the Northern Hemisphere through the
Bering Strait. This flora was subsequently broken up
by Pleistoceneglaciation, resulting in thedisjunction of
taxa on different continents. Gray’s hypothesis was
further developed by Chaney (1947, 1959) who pro-
posed the ‘‘Arcto-Tertiary Geoflora’’ concept to explain
the floristic similarities between eastern Asia and
North America. Chaney envisioned a common biota
F IG. 5.
names on a phylogenetic tree with the distributional area of the
respectivetaxon. WNA, western North America; ENA, eastern North
America; CA, Central America; and EA, eastern Asia.
Area cladograms derived by replacing terminal taxon
XIANG, SOLTIS, AND SOLTIS
with a taxonomic composition similar to that of the
modern North Temperate Flora occupied the high
latitudes of Eurasia and North America in the early
Tertiary. This Geoflora subsequently spread southward
with little changes in its composition. During the late
Tertiary and Quaternary, most of this flora became
extinct in Europe, western Asia, and central North
America, resulting in the disjunctions in the Northern
Hemisphere observed today. Updated paleontological
evidence, however, contradicts the Arcto-Tertiary Geo-
flora concept (Wolfe, 1969, 1972, 1975, 1985). The new
paleontological evidence suggests the presence of an
early Tertiary ‘‘boreotropic flora’’ with dynamic taxo-
nomic composition in theNorthern Hemisphere(Wolfe,
1969, 1972, 1975, 1985). Wolfe (1969, 1972, 1975, 1985;
see alsoTiffney, 1985a, 1985b) regarded the interconti-
nental similarities of the North Temperate Flora to be
the result of the spreading of the boreotropic flora. This
boreotropic flora developed into what has been termed
a ‘‘Mixed Mesophytic forest’’ (a warm-temperate type of
vegetation) in the mid-Tertiary (Wang, 1961; Tiffney,
1985a); this forest was once more or less continuously
distributed throughout the Northern Hemisphere, but
later climatic and geological changes resulted in its
range restriction and ultimately the disjunction ob-
served today (Li, 1952; Wood, 1971, 1972; Graham,
1972; Wolfe, 1972, 1981; Hsu, 1983; Tiffney, 1985a,
1985b). Thus, the extant eastern Asian-eastern and
western North American disjuncts are considered as
some of the remnants of this mixed mesophytic forest
(Li, 1952; Wood, 1971, 1972; Graham, 1972; Hsu, 1983).
Under this hypothesis, the pattern of phylogenetic
relationship observed suggests that the eastern and
western North American disjunction occurred after the
initial eastern Asian-North American isolation. Thus,
one of the first disruptions in the continuous Mixed
Mesophytic forest would have been the separation of
Eurasia from North America due to continental drift.
Various geologic and climatic changes later eliminated
this flora from much of Europe, central Asia, and
central North America (Graham, 1972; Leopold and
MacGinitie, 1972; Tiffney, 1985a, 1985b), resulting in
the disjunction in eastern Asia, eastern and western
Westress, however, that congruencebetween phylog-
enies and geographic distributions does not necessarily
indicate an identical phytogeographic history. The flo-
ristic disjunction in eastern Asia, eastern North
America, and western North America might have origi-
nated at very different geological times in different
genera, a phenomenon known as ‘‘pseudocongruence’’
(Cunningham and Collins, 1994), as pointed out by
Tiffney (1985a; see also Wolfe, 1969, 1972, 1981; Li,
1972). In this regard, information on timeof divergence
between species in different genera is critical to deter-
mine whether the observed identical phylogenetic pat-
tern among the diverse seven taxa represents pseudo-
congruenceor a shared floristic history.
Divergence time between lineages can be estimated
using fossil evidence or a molecular clock (although
difficulties with a molecular clock approach are well-
known; see below). All seven taxa discussed herein
except Cornus are either unknown as fossils (Boykinia,
Tiarella, Trautvetteria, and Adiantum pedatum) or
have an insufficient fossil record (Aralia sect. Aralia
and Calycanthus) to estimate the divergence times
between disjunct species (Collinson et al., 1993; Taylor
and Taylor, 1993; Friis et al, 1994; Lang, 1994; Mai,
1994). Fossil fruit stones of separate-fruited, big-
bracted dogwoods (like the extant North American
species) appear in the mid-Oligocene and Miocene
deposits in Europe, and fruit stones of compound-
fruited, big-bracted dogwoods (like the extant eastern
Asian species) werefoundin Europein thePlioceneand
in J apan in thePleistocene(seeEyde, 1988). This fossil
approximately 30–32 million years beforepresent (BP),
during the Oligocene; fossil evidence alsosuggests that
the eastern Asian and North American big-bracted
dogwoods diverged at least fivemillion years BP.
The molecular clock is based on the neutral theory of
molecular evolution and assumes a constant rate of
molecular evolution across lineages (Zuckerkandle and
Pauling, 1965; Kimura, 1983).Although thereis contro-
versy regarding the neutral theory, and heterogeneous
rates of molecular evolution have been documented in
different lineages (e.g., Wilson et al., 1990; Gaut et al.,
1992, 1993, 1996), a molecular clock may be useful for
estimating divergence times if the clock can be cali-
brated with some confidence. An overall divergence
rate of approximately 10?9nucleotide substitutions per
site per year has been estimated for the chloroplast
genome (Zurawski et al., 1984; Zurawski and Clegg,
1987). This molecular clock was used to estimate of
time of divergence between the eastern Asia-eastern
North America disjunct species pair in Liriodendron
(Magnoliaceae) based on cpDNA restriction site data
(Parks and Wendel, 1990). Theestimatewas congruent
with that from fossil evidence (11–14 million years BP)
(Parks and Wendel, 1990).
With the aim of gaining some insight intothe time of
divergence between species from eastern Asian and
both eastern and western North America, we applied
this molecular clock for three genera, Cornus, Boyki-
nia, and Calycanthus, that have cpDNA restriction site
data available.Assuming a sequence divergence rate of
0.1% per million years, the divergence time between
the eastern Asian and North American big-bracted
dogwoods is estimated to be 13.1 million years BP (in
the mid-Miocene), based on a 2.61% sequence diver-
gence estimated from restriction site mutations using
the computer program SDE 1.2 (Wolfe and Wolfe,
PHYLOGENETIC PATTERNS OF DISJ UNCT TAXA
1993) [The SDE program estimates sequence diver-
gence using the J ukes and Cantor (1969) method with
correction for multiple hits. Equations 5.3, 5.38, and
5.41 of Nei (1987), and 3.19, 3.32, and 3.33 of Li and
Graur (1991) were implemented in the calculation.]
This estimate falls in the time range estimated based
on fossil evidence of Cornus as discussed above. Diver-
gence time for the big-bracted dogwoods within the
NorthAmerican continent is estimated tobe9.7 million
years BP (in the mid-Miocene) based on an average of
1.94% sequencedivergencebetween species.
Using the same molecular clock and cpDNA restric-
tion site data, the divergence time between the eastern
Asian and North American species of Boykinia is
estimated to be only 2.6 million years BP (end of the
Pliocene) (0.51% sequence divergence between the two
lineages); between the eastern and western North
American species, the divergence time is 1.2 million
years BP (in the Pleistocene) (0.23% sequence diver-
gence). Similarly, for Calycanthus, the divergence time
between theeasternAsian and NorthAmerican species
is estimated at 3.1 million years BP, in the Pliocene
(0.61% sequence divergence), and that between the
eastern and western North American species is 2.6
million years BP, or the end of the Pliocene (0.52%
sequence divergence) (Wen et al., 1996). Divergence
times for Trautvetteria, Tiarella, and Aralia sect. Ara-
lia cannot be estimated using this same clock because
theoccurrenceof cpDNA captureinvolving Tiarella and
Heuchera would distort the estimate, and cpDNA re-
striction site data are not available for Trautvetteria
and Aralia sect. Aralia.
This analysis of divergence times suggests that spe-
cies of Cornus diverged much earlier than species of
Boykinia and Calycanthus, with species of Boykinia
diverging most recently. The different divergence times
estimated for species of these genera may indicate that
either the isolation of species in these genera occurred
at different geological times, or, alternatively, these
genera simply have experienced unequal rates of mo-
To distinguish between these two possibilities, rela-
tiveratetests of cpDNA evolution need tobeperformed
for Cornus, Boykinia, and Calycanthus. If the different
divergence times estimated for Cornus, Boykinia, and
Calycanthus weretotally attributabletotheratediffer-
ences, we would expect to see a much higher rate in
Cornus, but similar rates in Boykinia and Calycanthus.
Because assessment of homology of restriction site
mutations in distantly related taxa becomes problem-
atic, such tests are not feasible with restriction site
data for these three genera that belong to three differ-
ent subclasses of flowering plants (Asteridae, Rosidae,
andMagnoliidae, respectively). Wethereforeconducted
relativeratetests for thechloroplast generbcL in these
genera. The tests were performed for both synonymous
and nonsynonymous substitutions using the method of
Li and Bousquet (1992) and following Gaut et al. (1996)
and Xiang et al. (1998a). The rbcL sequence of Cerato-
phyllum, a genus that is thesister toall other flowering
plants in the rbcL sequence analysis of Chase et al.
(1993), was first used as the reference sequence for the
tests. Because Ceratophyllum has a relatively long
branch (44) in the rbcL tree of Chase et al. (1993), its
sister relationship to all other flowering plants may be
questionable. We therefore performed the tests using a
second reference sequence, the rbcL of Nymphaea, a
member of ‘‘Paleoherb II’’ in the rbcL analysis of Chase
et al. (1993), tocomparetheresults.
When Ceratophyllum was used as the reference
taxon, the rates of evolution of rbcL do not differ
significantly between Cornus and Boykinia for either
synonymous or nonsynonymous substitutions (Table
2), but dodiffer significantly between Calycanthus and
Boykinia for both synonymous and nonsynonymous
substitutions (Table 2). The rate of synonymous substi-
tutions in Calycanthus is lower than that in Boykinia,
but the rate of nonsynonymous substitutions is the
reverse (Table 2). A similar situation exists between
Cornus and Calycanthus (see Table 2). Cornus has a
higher synonymous substitution rate, but a lower non-
synonymous substitution rate compared to Calycan-
thus, suggesting that the total substitution rates be-
tween these genera may not be significantly different.
TABL E 2
R esults of R elative R ate Tests at rbcL for Cornus,
Boykinia, and Calycanthus
Reference taxon: Ceratophyllum
Reference taxon: Nymphaea
Note. Ceratophyllum demersum L. and Nymphaea odorata Aiton
were used as reference taxa for the tests. A, Cornus canadensis; B,
Calycanthus floridus; and C, Boykinia rotundifolia. The rbcL se-
quences of taxa used in the tests were from Xiang et al. (1993) for
Cornus, from Chase et al. (1993) for Ceratophyllum, Nymphaea, and
Calycanthus, and from Soltis et al. (1993) for Boykinia. K1–K2,
Difference in the weighted number of substitutions per site between
lineage 1 and lineage 2; Ks, synonymous substitutions; Ka, nonsyn-
onymous substitutions; and Kt, Total substitutions. For Ks and Ka, a
test statistic with absolute value ?1.96 (indicated in bold) is signifi-
cant at the 0.05 level. For Kt, a value of K1–K2 greater than two
times of the standard error is considered significant at the 0.05 level.
XIANG, SOLTIS, AND SOLTIS
Our subsequent tests for total substitution rates using
theKimura-2-parameter distancemethod of MEGA 1.0
(Kumar et al., 1993) following Li (1997, pages 218–219)
indeed showed no significant differences among the
three genera (Table 2). When Nymphaea was used as
the reference taxon, both synonymous and nonsynony-
moussubstitution ratesdonot differ significantly among
the three genera except for the nonsynonymous substi-
tution rate between Cornus and Boykinia, which differ
significantly (Table 2). These results apparently donot
fit the expectation that the longer divergence estimate
for Cornus is due to an elevated rate of rbcL evolution.
Instead, the different divergence times obtained for
Cornus, Boykinia, and Calycanthus may provide evi-
dencefor pseudocongruence; that is, thesethreegenera
obtained their disjunct distributions in eastern Asia,
eastern North America, and western North America at
different geological times.
Therefore, the seven diverse genera examined in this
study do not necessarily share an identical biogeo-
graphic history although they exhibit the same pattern
of phylogenetic relationship among species. Some gen-
era, such as the big-bracted dogwoods, may be relicts of
the Mixed Mesophytic forest and were once continu-
ously distributed in the Northern Hemisphere, given
that thespecies diverged near themiddleMiocene. The
big-bracted dogwoods probably obtained their distribu-
tion in the twocontinents via the Bering land bridge or
via a series of island ‘‘stepping stones’’ in the North
Atlantic (see Tiffney, 1985a, 1985b). Others, such as
Calycanthus (producing seeds enclosed in fleshy recep-
tacles that may be eaten by birds) and Boykinia (bear-
ing tuberculate seeds that may become caught in
feathers of birds), may have obtained their disjunct
distributions through long-distance dispersal, given
that species in these genera were isolated relatively
recently based on the molecular clock. Alternatively,
these two genera may have obtained their disjunct
distributions through gradual migration across the
Bering land bridge, followed by long-distance dispersal
within North America. The Bering land bridge was
periodically available for exchanges of plants between
eastern Asia and western North America almost
throughout the Tertiary (until 3.5 million years BP; see
Allen, 1983; Tiffney, 1985a, 1985b; Cunningham and
Collins, 1994). The North Atlantic bridge was, in con-
trast, functional until only the late Eocene, although a
series of island ‘‘stepping stones’’ existed in the North
Atlantic until the early Oligocene (Tiffney, 1985a).
Exchanges of elements between eastern and western
North American floras ceased in the Miocene (Tiffney,
1985a). Our molecular data thus donot refute the view
that the floristic similarities between eastern Asia and
North America have a complex history involving mul-
tiple historical events (both vicariant and dispersal) at
different geological times in different taxa (Wolfe, 1969,
1972, 1975,1981, 1985; Li, 1972; Tiffney, 1985a).
Center of origin.
consistent pattern of phylogenetic relationship among
six genera of flowering plants and a fern species found
in eastern Asia and both eastern and western North
America, our analyses suggest that the continent of
origin for these genera may differ or simply be ambigu-
ous. We preface this discussion by stressing the vagar-
ies and difficulties of inferring a center of origin based
on the mapping of geographic distribution ontophylog-
enies. For Calycanthus and Aralia sect. Aralia, the
continent of origin appears to be Asia. Conversely,
broad phylogenetic analyses of the Boykinia group, as
well as of the entire Saxifragaceae s. s. (Soltis et al.,
1993, 1995, 1996), indicate that for Boykinia, western
North America is the center of origin. For many of the
genera examined, however, the results are equivocal.
For example, thesister group of Trautvetteria is a clade
composed of Ranunculus and Myosurus, two genera
that are widely distributed in both the Northern and
Southern Hemispheres and for which phylogenetic
relationships among species are poorly understood.
Similarly, the sister group of the big-bracted dogwoods
has a circumboreal distribution; thus the center of
origin is equivocal. For Tiarella, no well-supported
sister group has yet been identified in several phyloge-
netic analyses; hence, no center of origin can yet be
Fossil evidence would permit further evaluation of
the biogeographic hypothesis on the center of origin
inferred from phylogenetic analyses of molecular data.
As mentioned above, none of the seven taxa discussed
herein except Cornus has an adequatefossil record.
Although we have identified a
Six genera of angiosperms, Boykinia, Tiarella,
Trautvetteria, the big-bracted dogwoods of Cornus,
Aralia sect. Aralia, and Calycanthus, and the fern
species, Adiantum pedatum, show a similar pattern of
biogeographic relationships: the eastern and western
North American species/populations are sister groups;
this clade is in turn the sister of the eastern Asian
species/populations. Under the principle of historical
biogeography (i.e., congruence between phylogenetic
topologies and geographic distribution of lineages indi-
cates shared biogeographic histories), the phylogenetic
data presented here lend support to the vicariant
hypothesis that this well-known floristic disjunction
represents the fragmentation of a more or less continu-
ous Mixed Mesophytic forest community that existed
during the Tertiary. We caution, however, that congru-
encebetween phylogenies and geographic distributions
does not necessarily indicate an identical phytogeo-
graphic history. Thefloristic disjunction involving east-
ern Asia, eastern North America, and western North
America may have originated at very different geologi-
cal times in different genera (e.g., Li, 1972; Tiffney,
PHYLOGENETIC PATTERNS OF DISJ UNCT TAXA
1985b), a phenomenon known as ‘‘pseudocongruence’’
(Cunningham and Collins, 1994). Our initial attempts
to evaluate this possibility using a molecular clock to
estimate divergence times donot refute the hypothesis
ofpseudocongruence.Other taxa displayingthisdisjunc-
tion should be examined for phylogenetic relationships
and time of divergence between disjunct species, and
additional efforts should be made to evaluate the
possibility of pseudocongruence.
We thank H. Ohba, H. Forbes, G.-H. Zhu, and S. Kawano for
assistance in the field or for providing plant material, S. Hoot for
kindly providing DNA samples of MyosurusandTrautvetteria carolin-
ensis, P. Crane for recommending references on fossils, and D. J .
Crawford and J . J . Doyle for reading the manuscript and providing
valuable comments. We have alsobenefited from discussions with D.
E. Boufford, R. Olmstead, S. Hoot, J . Wen, and C. A. Paris. This
project was supported by NSF Grants DEB92–24072 and DEB
Allen, R. T. (1983). Distribution patterns among Arthropods of the
north temperate deciduous forest biota. Ann. Missouri Bot. Gard.
Baldwin, B. G. (1992). Phylogenetic utility of theinternal transcribed
spacers of nuclear ribosomal DNA in plants: An example from the
Compositae. Mol. Phylogenet. Evol. 1: 3–16.
Boufford, D. E. (1992). Affinities in the floras of Taiwan and eastern
NorthAmerica. Phytogeogr. Bot. Invent. Taiwan, Monogr. 12: 1–16.
Boufford, D. E., and S. A. Spongberg. (1983). Eastern Asian-eastern
North American phytogeographical relationships—A history from
the time of Linnaeus to the twentieth century. Ann. Missouri Bot.
Gard. 70: 423–439.
Boufford, D. E., J . V. Crisci, H. Tobe, andP. C. Hoch. (1990).A cladistic
analysis of Circaea (Onagraceae). Cladistics 6: 171–182.
Bremer, K. (1988). The limits of amino acid sequence data in
angiosperm phylogenetic reconstruction. Evolution 42: 795–803.
Brown, J . H., and A. C. Gibson. (1983). ’’Biogeography.’’ Mosby, St.
Chaney, R. W. (1947). Tertiary centers and migration routes. Ecol.
Monogr. 17: 139–148.
Chaney, R. W. (1959). Miocene forests of the Columbia Plateau, I.
Composition and interpretation. Carnegie Inst. Wash. Publ. 617:
Chase, M. W., D. E. .Soltis, R. G. Olmstead, D. Morgan, D. H. Les,
B. D. Mishler, M. R. Duvall, R.A. Price, H. G. Hills, Y.-L. Qiu, K.A.
Kron, J . H. Rettig, E. Conti, J . D. Palmer, J . R. Manhart, K. J .
Sytsma, H. J . Michaels, W. D. Clark, M. Hedren, K.-J . Kim, C. F.
Wimpee, J . F. Smith, G. R. Furnier, S. H. Strauss, Q.-Y. Xiang,
G. M. Plunkett, P. S. Soltis, S. M. Swensen, S. E. Williams, P. A.
Gadek, C. J . Quinn, L. E. Eguiarte, E. Golenberg, G. H. Learn, J r.,
S. W. Graham, S. C. H. Barrett, S. Dayanandan, and V. A. Albert.
(1993). Phylogenetics of seed plants: An analysis of nucleotide
sequences from theplastid generbcL. Ann. Missouri Bot. Gard. 80:
Collinson, M. E., M. C. Boulter, and P. L. C. Holmes. (1993).
Magnoliophyta (‘‘Angiospermae’’). In ‘‘Fossil Record 2’’ (M. J .
Benton, Ed.), pp. 801–841. Chapman and Hall, London.
Cunningham, C. W., and T. M. Collins. (1994). Developing model
systems for molecular biogeography: vicarianceand interchangein
marine invertebrates. In ‘‘Molecular Ecology and Evolution, Ap-
proaches andApplication’’ (B. Schierwater, B. Streit, G. P. Wagner,
and R. DeSalle, Eds.), pp. 405–433. Birkhauser Verlad Base,
Eernisse, D. J ., and A. G. Kluge. (1993). Taxonomic congruence
versus total evidence, and amniotephylogeny inferred from fossils,
molecules, and morphology. Mol. Biol. Evol. 10: 1170–1195.
Eyde, R. H. (1985). The case for monkey-mediated evolution in
big-bracted dogwoods. Arnoldia (J amaica Plain) 45: 2–9.
Eyde, R. H. (1988). Comprehending Cornus: Puzzles and progress in
thesystematics of thedogwoods. Bot. Rev. 54: 233–351.
Felsenstein, J . (1985). Confidencelimits on phylogenies:An approach
using thebootstrap. Evolution 39: 783–791.
Friis, E. M., H. Eklund, K. R. Pedersen, and P. R. Crane. (1994).
Virginianthus calycanthoides gen. et sp. nov.—A callycanthaceous
flower from the Potomac group (early Cretaceous) of eastern North
America. Int. J . Pl. Sci. 155: 772–785.
Futuyma, D. J . (1986). ’’Evolutionary Biology,’’ 2nd. ed., Sinauer,
Gaut, B. S., S. V. Muse, W. D. Clark, and M. T. Clegg. (1992). Relative
rates of nucleotidesubstitution at therbcL locus of monocotyledon-
ous plants. J . Mol. Evol. 35: 292–303.
Gaut, B. S., S. V. Muse, and M. T. Clegg. (1993). Relative rates of
nucleotide substitution in the chloroplast genome. Mol. Phyl. Evol.
Gaut, B. Morton, B. C. McCaig, and M. T. Clegg. (1996). Substitution
rate comparisons between grasses and palms: synonymous rate
differences at the nuclear gene Adh parallel rate differences at the
plastid generbcL. Proc. Natl. Acad. Sci. USA 93: 10274–10279.
Gornall, R. J ., and B. A. Bohm. (1985). A monograph of Boykinia,
Peltoboykinia, Bolandra and Suksdorfia (Saxifragaceae). Bot. J .
Linn. Soc. 90: 1–71.
Graham, A. (1972). Outline of the origin and historical recognition of
floristic affinities between Asia and eastern North America. In
‘‘Floristics and Paleofloristics of Asia and Eastern North America’’
(A. Graham, Ed.), pp. 1–18. Elsevier, Amsterdam.
Gray, A. (1859). Diagnostic characters of new species of phaneroga-
mous plants collected in J apan by Charles Wright, Botanists of the
U. S. North Pacific Exploring Expedition, with observations upon
the relationship of the J apanese flora tothat of North America and
of other parts of the northern Temperate zone. Mem. Am. Acad.
Arts 6: 377–453.
Gray, A. (1878). Forest geography and archaeology. A lecture deliv-
eredbeforetheHarvardUniversity Natural History Society. Am. J .
Sci. Arts Ser. 3 16: 85–94, 183–196.
Hara, H. (1972). Patterns of differentiation in flowering plants. In
‘‘Floristics and Paleofloristics of Asia and Eastern North America’’
(A. Graham, Ed.), pp. 50–60. Elsevier, Amsterdam.
Hong, D.-Y. (1983). The distribution of Scrophulariaceae in the
Holarctic with special reference to the floristic relationships be-
tween eastern Asia and eastern North America. Ann. Missouri Bot.
Gard. 70: 701–712.
Hong, D.-Y. (1993). Eastern Asian–North American disjunctions and
their biological significance. Cathaya 5: 1–39.
Hoot, S. B. (1995). Phylogeny of theRanunculaceaebased on prelimi-
nary atpB, rbcL and 18S nuclear ribosomal DNA sequencedata. Pl.
Syst. Evol. [Suppl] 9: 241–251.
Hsu, J . (1983). Late Cretaceous and Cenozoic vegetation in China,
emphasizing their connections with North America. Ann. Missouri
Bot. Gard. 70: 490–508.
Hu, H. H. (1935). A comparison of the ligneous flora of China and
eastern North America. Bull. ChineseBot. Soc. 1: 79–97.
Humphries, C. J ., and L. R. Parenti. (1986). ’’CladisticBiogeography.’’
Clarendon Press, Oxford, England.
XIANG, SOLTIS, AND SOLTIS
Iltis, H. H. (1983). Biogeographical relationships between temperate
eastern Asia and temperate eastern North America. In ‘‘Proceed-
ings of the U. S.—J apan Science Seminar on the Origin of the East
Asian and North American Floras from Arctotertiary Precursors’’
(UnitedState-J apan CooperativeScienceProgram), pp. 49–51. The
New York Botanical Garden Cary Arboretum, Millbrook.
J ohnson, L. A., and D. E. Soltis. (1994). matK DNA sequences and
phylogenetic reconstruction in Saxifragaceae sensu stricto. Syst.
Bot. 19: 143–156.
J ohnson, L. A., and D. E. Soltis. (1995). Phylogenetic inference in
Saxifragaceae sensu strictoand Gilia (Polemoniaceae) using matK
sequences. Ann. Missouri Bot. Gard. 82: 149–175.
J ukes, T. H., and C. R. Cantor. (1969). Evolution of protein molecules.
In ‘‘Mammalian Protein Metabolism’’ (H. N. Munro, Ed.), pp.
Pages21–132.Academic Press, New York, NY.
Kato, H., S. Kawano, R. Terauchi, M. Ohara, and F. H. Utech. (1995).
Evolutionary biology of Trillium and related genera (Trilliaceae): I.
Restriction site mapping and variation of chloroplast DNA and its
systematic implications. Pl. Sp. Biol. 10: 17–29.
Kimura, M. (1983). ’’The Neutral Theory of Molecular Evolution.’’
CambridgeUniv. Press, Cambridge, England.
Koyama, T., and S. Kawano. (1964). Critical taxa of grasses with
North American and eastern Asiatic distribution. Can. J . Bot. 42:
Koyama, T. (1983). Cyperaceae with eastern Asiatic and North
American disjunct distribution. In ‘‘Proceedings of the U. S.—
J apan Science Seminar on the Origin of the East Asian and North
American Floras from Arctotertiary Precursors’’ (United State-
J apan Cooperative Science Program), p. 43. The New York Botani-
cal Garden, Cary Arboretum, Millbrook.
Kumar, S., K. Tamura, and M. Nei. (1993). ’’MEGA: Molecular
Evolutionary Genetic Analysis,’’ Version 1.0. Pennsylvania State
University, University Park, PA.
Lang, G. (1994). ’’Quartare Vegetationsgeschichte Europas—Meth-
oden und Ergebnisse.’’ Gustav Fischer Verlag, New York.
Leopold, E. B., and H. D. MacGinitie. (1972). Development and
affinities of Tertiary floras in the Rocky Mountains. In ‘‘Floristics
andPaleofloristics ofAsia andEastern NorthAmerica’’(A. Graham,
Ed.), pp. 147–200. Elsevier, Amsterdam.
Li, H. L. (1952). Floristic relationships between eastern Asia and
eastern North America. Trans. Am. Philos. Soc. 42: 371–429.
Li, H. L. (1972). EasternAsia–eastern NorthAmerica species-pairs in
wide-ranging genera. In ‘‘Floristics and Paleofloristics of Asia and
Eastern North America’’ (A. Graham, Ed.), pp. 65–78. Elsevier,
Li, P., and J . Bousquet. (1992). Relative rate test for nucleotide
substitutions between two lineages. Mol. Biol. Evol. 9: 1185–1189.
Li, W. H. (1997). ’’Molecular Evolution,’’ pp. 215–219. Sinauer,
Li, W. H., and D. Graur. (1991). ’’Fundamentals of Molecular Evolu-
tion.’’ Sinauer, Sunderland, MA.
Maddison, W. P., and D. R. Maddison. (1992). ’’MacClade: Interactive
Analysis of Phylogeny and Character Evolution,’’ Version 3.0.
Sinauer Associates, Sunderland, MA.
Mai, D. H. (1994). ’’Tertiare Vegetationsgeschichte E uropas—
Methoden und Ergebnisse.’’ Gustav Fischer Verlag, New York.
Morgan, D. R., and D. E. Soltis. (1993). Phylogenetic relationships
among members of Saxifragaceae sensu lato based on rbcL se-
quencedata. Ann. Missouri Bot. Gard. 80: 631–660.
Murrell, E. Z. (1993). Phylogenetic relationships in Cornus (Corna-
ceae). Syst. Bot. 18: 469–495.
Nei, M. (1987). ’’Molecular Evolutionary Genetics.’’ Columbia Univ.
Press, New York, NY.
Nelson, G., and N. Platnick. (1981). ’’Systematics and Biogeography:
Cladistics and Vicariance.’’ Columbia Univ. Press, New York, NY.
Oosterbroek, P., and J . W. Arntzen. (1992). Area-cladograms of
Circum-Mediterranean taxa in relation to Mediterranean palaeo-
geography. J . Biogeogr. 19: 3–20.
Paris, C.A. (1991). ’’Molecular Systematics of theAdiantum pedatum
Complex: Phylogeny and Biogeography,’’ Ph. D. dissertation, Uni-
versity of Vermont, Burlington, VT.
Paris, C. A., and C. H. Haufler. (1994). Geographic isolation and
allopatricspeciation in ferns with a disjunct distribution in eastern
Asia and eastern North America. Am. J . Bot. 81: 137. [Abstract]
Parks, C. R., and J . F. Wendel. (1990). Molecular divergence between
Asian and North American species of Liriodendron (Magnoliaceae)
with implications for interpretation of fossil floras. Am. J . Bot. 77:
Platnick, N., and G. Nelson. (1978). A method of analysis for
historical biogeography. Syst. Zool. 27: 159–188.
Qiu, Y.-L., M. W. Chase, D. H. Les, and C. R. Parks. 1993. Molecular
phylogenetics of the Magnoliidae: Cladistic analyses of nucleotide
sequences of the plastid gene rbcL. Ann. Missouri Bot. Gard. 80:
Sober, E. (1988). The conceptual relationship of cladistic phylogenet-
ics and vicariancebiogeography. Syst. Zool. 37: 245–253.
Soltis, D. E., and B. A. Bohm. (1984). Karyology and flavonoid
chemistry of the disjunct species of Tiarella (Saxifragaceae). Syst.
Bot. 9: 441–447.
Soltis, D. E., and R. Kuzoff. (1995). Discordance between nuclear and
chloroplast phylogenies in the Heuchera Group (Saxifragaceae).
Evolution 49: 727–742.
Soltis, D. E., P. S. Soltis, T. G. Collier, and M. L. Edgerton. (1991).
Chloroplast DNA variation within and among genera of the
Heuchera group (Saxifragaceae): Evidence for chloroplast transfer
and paraphyly. Am. J . Bot. 78: 1091–1112.
Soltis, D. E., D. R. Morgan, A. Grable, P. S. Soltis, and R. Kuzoff.
(1993). Molecular systematics of Saxifragaceae sensu stricto. Am.
J . Bot. 80: 1056–1081.
Soltis, D. S., Q-Y. Xiang, and L. Hufford. (1995). Relationships and
evolution of Hydrangeaceae based on rbcL sequence data. Am. J .
Bot. 82: 504–514.
Soltis, D. E., L. A. J ohnson, and C. Looney. (1996). Discordance
between ITS and chloroplast topologies in the Boykinia group
(Saxifragaceae). Syst. Bot. 21: 169–186.
Swofford, D. L. (1993). ’’PAUP: Phylogenetic Analysis Using Parsi-
mony,’’ Version 3.1.1. Illinois Natural History Survey, Champaign.
Tamura, M. (1983). Distribution and differentiation of someranuncu-
laceous genera in East Asia and North America. In ‘‘Proceedings of
the U. S. – J apan Science Seminar on the Origin of the East Asian
and NorthAmerican Floras fromArctotertiary Precursors’’ (United
State-J apan Cooperative Science Program), pp. 9–18. The New
York Botanical Garden, Cary Arboretum, Millbrook.
Taylor, T. N., and E. L. Taylor. (1993). ’’The Biology and Evolution of
Fossil Plants,’’ p. 430. PrenticeHall, Englewood Cliffs, NJ .
Tiffney, B. H. (1985a). The Eocene North Atlantic land bridge: Its
importance in Tertiary and modern phytogeography of the North-
ern Hemisphere. J . Arn. Arb. 66: 243–273.
Tiffney, B. H. (1985b). Perspectives on the origin of the floristic
similarity between eastern Asia and eastern North America. J .
Arn. Arb. 66: 73–94.
Wang, C.-W. (1961). The forests of China. Harvard Univ., Maria
Moors Cabot Found., Publ. 5: 1–313.
Wen, J ., and T. F. Stuessy. (1993). Thephylogeny and biogeography of
Nyssa (Cornaceae). Syst. Bot. 18: 68–79.
Wen, J ., R. K. J ansen, and E. A. Zimmer. (1996). Phylogenetic
relationships and DNA sequence divergence of eastern Asian and
PHYLOGENETIC PATTERNS OF DISJ UNCT TAXA
eastern North American disjunct plants. In ‘‘Current Topics on
Molecular Evolution’’ (M. Nei, and N. Takahata, Eds.), pp. 37–44.
Pennsylvania State University, U.S.A., and The Graduate Univer-
sity for AdvanceStudies, Hayama, J apan.
Wilson, M. A., B. Gaut, and M. T. Clegg. (1990). Chloroplast DNA
evolves slowly in thepalm family. Mol. Biol. Evol. 7: 303–314.
Wolfe, J . A. (1969). Neogene floristic and vegetational history of the
Pacific Northwest. Madrono20: 83–110.
Wolfe, J . A. (1972). An interpretation of Alaskan Tertiary floras. In
‘‘Floristics and Paleofloristics of Asia and Eastern North America’’
(A. Graham, Ed.), pp. 201–233. Elsevier, Amsterdam.
Wolfe, J . A. (1975). Some aspects of plant geography of the Northern
Hemisphere during the late Cretaceous and Tertiary. Ann. Mis-
souri Bot. Gard. 62: 264–279.
Wolfe, J . A. (1981). Vicariance biogeography of angiosperms in
relation to paleobotanical data. In ‘‘Vicariance Biogeography, A
Critique’’ (G. Nelson, and E. E. Rosen, Eds.), pp. 413–427. Colum-
bia Univ. Press, New York, NY.
Wolfe, J . A. (1985). Distribution and major vegetation types during
the Tertiary. In ‘‘The Carbon Cycle and the Atmospheric CO2:
Natural VariationsArchean toPresent’’ (E. T. Sundquist, and W. S.
Broecker, Eds.), pp. 357–375. Vol. 32.American Geophysical Union,
Wolfe, J . A., and E. B. Leopold. (1967). Neogene and Early Quater-
nary vegetation of northwestern North America and northeatern
Asia. In ‘‘The Bering Land Bridge’’ (D. M. Hopkiins, ed.), pp.193–
206. Stanford Univ. Press, Stanford, CA.
Wolfe, S. A., and A. D. Wolfe. (1993). SDE: Sequence divergence
estimator, Version 1.2. A computer program distributed by the
authors. OhioStateUniv., Columbus, OH.
Wood, C. E. (1971). Somefloristic relationships between thesouthern
Appalachians and western North America. In ‘‘The Distributional
History of the Bbiota of the Southern Appalachians, Part II. Flora’’
(P. C. Holt, Ed.), pp. 331–404. Virginia Polytechnic Institute and
StateUniv., Blacksburg, VI.
Wood, C. E. (1972). Morphology and phytogeography: The classical
approach tothe study of disjunctions. Ann. Missouri Bot. Gard. 59:
Wu, Z. Y. (1983). On the significance of Pacific intercontinental
discontinuity. Ann. Missouri Bot. Gard. 70: 577–590.
Xiang, Q.-Y. (1987). System and synopsis of Cornus subgen. Syn-
carpea (Nakai) Q. Y. Xiang (Cornaceae). Bull. Bot. Res. (Harbin)
Xiang, Q.-Y. (1995). ’’Molecular Systematics and Biogeography of
Cornus L. and PutativeRelatives,’’Ph. D. dissertation. Washington
StateUniv., Pullman, WA.
Xiang, Q.-Y., D. E. Soltis, D. R. Morgan, and P. S. Soltis. (1993).
Phylogenetic relationships of Cornus L. sensu lato and putative
relatives inferred from rbcL sequence data. Ann. Missouri Bot.
Gard. 80: 723–734.
Xiang, Q.-Y., S. J . Brunsfeld, D. E. Soltis, and P. S. Soltis. (1996).
Phylogenetic relationships in Cornus based on chloroplast DNA
restriction sites: Implications for biogeography and character
evolution. Syst. Bot. 21: 515–534.
Xiang, Q.-Y., and D. E. Soltis. (In press). rbcL sequence divergence
and phylogenetic relationships of Cornaceae sensu lato. In ‘‘Sino-
J apanese Flora - Its Characteristics and Diversification’’ (D. E.
Boufford, and H. Ohba, Eds.). TheUniv. TokyoPress, Tokyo.
Xiang, Q.-Y., D. E. Soltis, and P. S. Soltis. (1998a). Phylogenetic
relationships of Cornaceae and close relatives inferred from matK
and combined rbcL-matK sequences. Am. J . Bot. 85: 135–143.
Xiang, Q.-Y., D. J . Crawford, A. D. Wolfe, Y.-C. Tang, and C. W.
DePamphilis. (1998b). Origin and biogeography of Aesculus L.
(Hippocastanaceae) in a molecular phylogeneticperspective.Evolu-
Zuckerkandle, E., and L. Pauling. (1965). Evolutionary divergence
and convergence in proteins. In ‘‘Evolving Genes and Proteins’’ (V.
Bryson, and H. J . Vogel, Eds.), pp. 97–166. Academic Press, New
Zurawski, G., M. T. Clegg, and A. H. D. Brown. (1984). The nature of
nucleotide sequence divergence between barley and maize chloro-
plast DNA. Genetics 106: 735–749.
Zurawski, G. T., and M. T. . (1987). Evolution of higher-plant
chloroplast DNA-encoded genes: Implications for structure-
function and phylogenetic studies. Ann. Rev. Pl. Physio. 38:
XIANG, SOLTIS, AND SOLTIS