Molecular Phylogenetics and Evolution 52 (2009) 806–824
Contents lists available at ScienceDirect
Molecular Phylogenetics and Evolution
journal homepage: www.elsevier.com/locate/ympev
Rapid radiation of Impatiens (Balsaminaceae) during Pliocene and Pleistocene:
Result of a global climate change
Steven B. Janssens a,*, Eric B. Knox b, Suzy Huysmans a, Erik F. Smets a,c, Vincent S.F.T. Merckx a
a
Laboratory of Plant Systematics, Institute of Botany and Microbiology, K.U.Leuven, Kasteelpark Arenberg 31, P.O. Box 2437, BE-3001 Leuven, Belgium
Department of Biology, Indiana University, Jordan Hall 142, Bloomington, IN 47405, USA
c
National Herbarium of the Netherlands, University Leiden Branch, NL-2300 RA Leiden, The Netherlands
b
a r t i c l e
i n f o
Article history:
Received 28 November 2008
Revised 3 April 2009
Accepted 14 April 2009
Available online 3 May 2009
Keywords:
Global cooling
Impatiens
Pliocene
Pleistocene
Rapid radiation
a b s t r a c t
Impatiens comprises more than 1000 species and is one of the largest genera of flowering plants. The
genus has a subcosmopolitan distribution, yet most of its evolutionary history is unknown. Diversification analyses, divergence time estimates and historical biogeography, illustrated that the extant species
of Impatiens originated in Southwest China and started to diversify in the Early Miocene. Until the Early
Pliocene, the net diversification rate within the genus was fairly slow. Since that time, however, approximately 80% of all Impatiens lineages have originated. This period of rapid diversification coincides with
the global cooling of the Earth’s climate and subsequent glacial oscillations. Without this accelerated
diversification rate, Impatiens would only have contained 1/5th of its current number of species, thereby
indicating the rapid radiation of the genus.
Ó 2009 Elsevier Inc. All rights reserved.
1. Introduction
Global climate changes have played a crucial role in shaping
species evolution either by increased rates of extinction, or by
bursts of diversification. Earth’s climate has changed dramatically
altered during the past 65 million years and likewise several lines
of evidence have been found that correlate these fluctuations with
an increase in speciation rate (Bowie et al., 2004; Erkens et al.,
2007; Gamble et al., 2008; Hughes and Eastwood, 2006; Jaramillo
et al., 2006; Johnson et al., 2006; Merckx et al., 2008; Plana et al.,
2004; Richardson et al., 2001). Nevertheless, it is still being debated when the majority of the current angiosperm diversity has
originated during the course of evolution (Davies et al., 2004).
For a long time, it was considered that the recent climatic oscillations of the Pleistocene and Pliocene were crucial in promoting
general diversification among tropical and subtropical species
(Aubréville, 1962; Crowe and Crowe, 1982; Grey-Wilson, 1980a;
Haffer, 1969; Mayr and O’Hara, 1986; Moritz et al., 2000; Prance,
1973). This hypothesis of recent speciation in the Quaternary is
also referred to as the ‘refuge’ hypothesis and assumes allopatric
speciation in populations as they became isolated from each other
due to glacial aridity and rainforest fragmentation. Despite possible molecular evidence for recent explosive diversification events
in some tropical lineages (Harris et al., 2000 on Aframomum; Rich* Corresponding author. Fax: +32 16 32 19 55.
E-mail address: Steven.Janssens@bio.kuleuven.be (S.B. Janssens).
1055-7903/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved.
doi:10.1016/j.ympev.2009.04.013
ardson et al., 2001 on Inga), several recent studies suggest that
these fluctuations played a much smaller role than previously
thought and additionally suggest that the massive divergence of
tropical species probably started already in the Tertiary or even
earlier (Bowie et al., 2004; Glor et al., 2001; Knapp and Mallet,
2003; Moritz et al., 2000; Pennington et al., 2004; Plana et al.,
2004; Zink and Slowinsky, 1995).
The Pleistocene refuge hypothesis invokes the occurrence of a
heterogeneous topography as cool and dry climate conditions
forced rainforest taxa to withdraw into small refugial pockets (Haffer, 1969; Haffer and Prance, 2001; Mayr and O’Hara, 1986; Prance,
1973). In Africa for instance most of the postulated distribution
areas of glacial forest refugia are assumed to be associated with
more mountainous areas (Maley, 1987; Prance, 1982). In South
America, the situation is more complex with several refuge areas
being located in lowland Amazonia, yet also some uplifted areas
are considered as possible refuges for the neotropical flora (Haffer,
1969, 1982; Prance, 1982). Moreover, endemic taxa are frequently
associated with montane regions as geological and altitudinal differences create many habitat types (Grey-Wilson, 1980a; Plana
et al., 2004). Investigating tropical or subtropical taxa which are often endemic to lowland or montane rainforests would make it possible to test the influence of glacial oscillations in light of the
current species diversity in the tropics. In this aspect, balsams
are of particular interest. Impatiens (Balsaminaceae) is among the
most species-rich genera of flowering plants with over 1000 species and several newly described species every year (e.g. Janssens
S.B. Janssens et al. / Molecular Phylogenetics and Evolution 52 (2009) 806–824
et al., in 2009; Pòcs, 2007). Diversity hotspots for the genus can be
observed in tropical Africa, the Himalayan region, Madagascar,
South India and Sri Lanka and Southeast Asia (Bhaskar and Razi,
1981; Chen, 2001; Grey-Wilson, 1980a, 1989; Rahelivololona
et al., 2003; Shimizu, 1979; Toppin, 1920). Balsams are a characteristic element of tropical and subtropical montane forests above
500–800 m, even sometimes growing as high as 5000 m. Although
there are notable exceptions, most species of Impatiens cannot endure persistent drought or extended exposure to direct sunlight
(Fischer, 2004). As a result Impatiens species are typically confined
to stream margins, waterside boulders, and wet and/or montane
forests. For moisture-loving plants like Impatiens, upland rainforests are islands of suitable habitats in a vast landscape that is at
least seasonally dry. By migrating along the mountain slopes with
its moist habitat, Impatiens is able to minimize the effect of large
climatic fluctuations. As a consequence, populations of Impatiens
are easily isolated from other closely related lineages, which could
result in an increase in the rate of species diversification.
The present paper investigates the origin and evolution of the
Balsaminaceae with special emphasis on tempo and timing of
diversification in the genus Impatiens. We assess the evolutionary
relationships between the major clades of Impatiens and date their
origin in order to gain insight into causes of speciation. In addition,
we examine whether the global cooling at the Miocene–Pliocene
boundary and the subsequent climatic fluctuations during Pliocene
and Pleistocene might have triggered adaptive radiation or
whether the vast number of Impatiens lineages had already been
largely established in the Tertiary. Furthermore, we provide additional understanding in the importance of the refuge theory as a
possible explanation for the present tropical biodiversity.
2. Materials and methods
2.1. Taxon sampling
Species of Impatiens were selected to represent the geographic
and taxonomic diversity of the genus. In total 117 Impatiens species
were included from which chloroplast (atpB-rbcL) and nuclear (introns 4 and 5 of ImpDEF1 and ImpDEF2) sequences have been produced. Janssens et al. (2006) demonstrated that Impatiens omeiana
is sister to the other sampled species of Impatiens. Therefore, this
species was chosen as outgroup for phylogenetic analyses of the
remaining samples. A list of taxa with authorities, localities, voucher
information and GenBank Accession Nos. is listed in Tables 1 and 2.
2.2. Molecular protocols and phylogenetic analysis
Total genomic DNA was extracted using a modified version of
the hot CTAB protocol (Janssens et al., 2006). Primers and temperature profiles used for the amplification of atpB-rbcL, ImpDEF1 and
ImpDEF2 follow Janssens et al. (2006) and Janssens et al. (2007),
respectively. Amplification reactions were carried out on a GeneAmp PCR system 9700 (Applied Biosystems). Cycle sequencing
reactions were performed as in Janssens et al. (2006).
Initial alignment of the DNA sequences was carried out with
CLUSTALX, followed by manual adjustment using MacClade 4.05
(Maddison and Maddison, 2002). The chloroplast atpB-rbcL and
the nuclear ImpDEF1/ImpDEF2 datasets were analyzed separately
and combined using a maximum likelihood (ML) based method
with RaxML (Stamatakis et al., 2005). A partition homogeneity test
(implemented in PAUP4.0b10a; Swofford, 2002) was used to evaluate whether the data matrices provide different signal in the combined analyses. The most appropriate substitution model for each
gene marker was determined using a series of likelihood ratio tests
as implemented in ModelTest 3.06 (Posada and Crandall, 1998).
807
ML analyses were carried out on the CIPRES computer cluster
(CyberInfrastructure for Phylogenetic RESearch) at the San Diego
Supercomputing Center (www.phylo.org). Model parameters were
computed with RAxML-VI HPC (Stamatakis, 2006). Two runs were
performed, starting from a completely randomized tree, each with
100 inferences on the original alignment. Non-parametric bootstrapping (ML-BS) was carried out by calculating six times 167 replicates on the CIPRES computer cluster. Because a few nodes in the
ML phylogram were collapsed and a fully resolved tree was necessary to calculate diversification rates and lineages-through-time
plots, we set the branch length of these nodes collapsed to
0.000001, under the assumption that a lack of resolution indicated
a rapid radiation (Harris et al., 2000; Hughes and Eastwood, 2006).
2.3. Molecular dating
When estimating the age of the most recently diversified lineages in Impatiens, we encountered several problems which made
it impossible to immediately calibrate the obtained Impatiens phylogeny. (1) Lack of a suitable fossil record for Impatiens made it
impossible to directly calibrate the Impatiens topology. Although
there is one fossil pollen record available for the genus (Impatiensiditis brevicolpus – Pliocene; Song et al., 2004) the assignment of this
fossil pollen to clades in Impatiens was impossible. (2) Lack of resolution with the chloroplast atpB-rbcL spacer for the most recently
diversified lineages. This issue could be solved by using the fastevolving introns 4 and 5 of ImpDEF1 and ImpDEF2, which are very
effective to elucidate relationships at inter- and intraspecific levels
(Janssens et al., 2007). However, because of their high nucleotide
and indel substitution rate, it was impossible to amplify and sequence these loci for other taxa than Impatiens. Due to the high
amount of missing data for ImpDEF1 and ImpDEF2, it was impossible
to use these loci for the construction of the asterid phylogeny, which
was applied estimate the age of the lineages towards Impatiens.
As a result, we followed a two-stage strategy to calculate the
divergence times for the most diversified branches in Impatiens.
By extending the atpB-rbcL dataset of Impatiens with 319 atpB-rbcL
accessions of asterid taxa from GenBank, we were able to cover almost all major lineages of this large subclass (Cornales – Cornaceae
and Hydrangeaceae; Ericales – Actinidiaceae, Ebenaceae, Ericaceae,
Lecythidaceae, Marcgraviaceae, Primulaceae, Sapotaceae, Styracaceae and Tetrameristaceae; Garryales – Garryaceae; Aquifoliales –
Aquifoliaceae and Helwingiaceae; Gentianales – Apocynaceae, Gentianaceae, Loganiaceae and Rubiaceae; Lamiales – Gesneriaceae,
Globulariaceae, Lamiaceae, Lentibulariacaeae, Orobanchaceae,
Plantaginaceae and Scrophulariaceae; Solanales – Solanaceae;
Apiales – Apiaceae and Araliaceae; Asterales – Asteraceae, Campanulaceae and Menyanthaceae; Dipsacales – Adoxaceae, Caprifoliaceae, Dipsacaceae and Valerianaceae). Two accessions from the
rosid clade were used as outgroup. This large-scale approach allowed us to integrate multiple fossil calibration points in order to
minimize the bias produced by a single calibration point (Merckx
et al., 2008). The asterid dataset contains 437 taxa and 3880 characters and was analyzed with RAxML using similar settings as for the
combined chloroplast-nuclear dataset of Impatiens. ML bootstrapping of the asterid dataset was also carried out as described above.
A v2 likelihood ratio test, used to assess rate heterogeneity among
lineages, indicated that the substitution rates in atpB-rbcL are not
clocklike. As a result, we applied a relaxed clock model using the
Penalized Likelihood method (PL; Sanderson, 2002) as implemented the r8s software package. The rooted ML phylogram was
used as input file for the dating analysis. The following calibration
points were used for age estimation: (1) the crown group of Diospyros, (Ebenaceae) constrained to a minimum of 54.9 mya corresponding to the Eocene age estimate of the fossil Austrodiospyros
cryptostoma (Basinger and Christophel, 1985); (2) minimum age
808
S.B. Janssens et al. / Molecular Phylogenetics and Evolution 52 (2009) 806–824
Table 1
Accession Nos., voucher data and origin of plant material for taxa included in the combined DNA analyses of Impatiens.
Taxon
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
aquatilis Hook.f.
arachnoides H.Perrier
assurgens Baker f.
aurea Muhl.
aureliana Hook.f.
auricoma Baill.
balfourii Hook.f.
balsamina L.
bequaertii De Wild.
begonifolia S. Akiyama and H. Ohba
bicaudata H. Perrier
bombycina Lobin and E. Fischer
briartii De Wild. and Th. Dur.
bururiensis Grey-Wilson
burtonii Hook.f. var. burtonii
campanulata Wight
capensis Meerb.
catati Baill.
cecili N.E.Br.
chevalieri Tardieu
Impatiens chinensis L.
Impatiens chungtienensis Y.L. Chen
Impatiens clavicornu Turcz.
Impatiens columbaria J.J. Bos
Impatiens conchibracteata Y.L. Chen
Impatiens confusa Grey-Wilson ssp. longicornu
Grey-Wilson.
Impatiens congolensis G.M. Schulze and R. Wilczek
Impatiens cuspidata Wight and Arn.
Impatiens cyanantha Hook.f.
Impatiens davidi Franch.
Impatiens desmantha Hook.f.
Impatiens digitata Warb. ssp. digitata
Impatiens eberhardtii Tardieu
Impatiens edgeworthii Hook.f.
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
engleri Gilg ssp. engleri
eryaleia Launert ssp. eryaleia
ethiopica Grey-Wilson
faberi Hook.f.
fenghwaiana Y.L. Chen
fischeri Warb.
Impatiens flaccida Arn.
Impatiens flanaganae Hemsl.
Impatiens forrestii Hook.f. ex W.W. Smith
Impatiens furcata H. Perrier
Impatiens glandulifera Arn.
Impatiens gordoni Horne
Impatiens hamata Warb.
Impatiens hawkeri W. Bull
Impatiens hians Hook.f.
Impatiens henslowiana Arn.
Impatiens hochstetteri Warb. ssp. hochstetteri
Impatiens hochstetteri Warb. ssp. hochstetteri
Impatiens hoehnelii T.C.E. Fr.
Impatiens hydrogetonoides Launert ssp.
hydrogetonoides
Impatiens imbecilla Hook.f.
Impatiens inaperta H. Perrier
Impatiens ioides G.M. Schulze
Impatiens irvingii Hook.f.
Impatiens kamerunensis Warb. ssp. kamerunensis
Impatiens kamerunensis Warb. ssp. obanensis (Keay)
Grey-Wilson
Impatiens keilii Gilg ssp. keilii
Impatiens keilii Gilg ssp. pubescens Grey-Wilson
Impatiens kerriae Craib
Origin
Voucher
Accession
No.
ImpDEF1
Accession
No.
ImpDEF2
Accession
No.
atpB-rbcL
China, Yunnan
Madagascar
Zambia
North American origin, cult. Holden arboretum
China, Yunnan
Comoros origin, cult. Bot. Gard. Marburg
Himalayan origin, cult. Denver Bot. Gard.
Indian origin, cult. Kruidtuin Leuven
Congo origin, cult. Bot. Gard. Koblenz Univ.
China, Yunnan
Madagascan origin, cult. by Ray Morgan, UK
African origin, cult. Bot. Gard. Koblenz Univ.
Zambia
Burundi
Uganda
South Indian origin, cult. by Ray Morgan, UK
North American origin, cult. Holden arboretum
Madagascar
Zimbabwe
Vietnam, Balat
Song CNY017 (NEU)
Janssens SJ010 (LV)
Dessein SD720 (BR)
Janssens SJ008 (LV)
Yuan CN2k1-56 (NEU)
Janssens SJ001 (LV)
33051 (DBG)
Janssens SJ003 (LV)
Fischer NE1 (NEU)
Yuan CN2k1-51 (NEU)
Ray Morgan s.n. (LV)
Fischer NE3 (NEU)
Dessein 1018 (BR)
Reekmans 8110 (BR)
Knox 2803 (LV)
Ray Morgan s.n. (LV)
Janssens SJ009 (LV)
Janssens SJ011 (LV)
Knox 4353 (LV)
Song and Puoong 2004–07
(NEU)
Yuan CN2k1-49 (NEU)
Yuan CN2k2-204 (NEU)
Ray Morgan s.n. (LV)
FB/S2966 (BR)
Hao 427 (NEU)
Knox 3239 (LV)
–
FJ826680
FJ826681
EF133560
EF133561
EF133562
EF133564
EF133563
FJ826682
EF133565
EF133566
FJ826683
FJ826684
FJ826685
EF133567
EF133568
FJ826686
FJ826687
–
EF133611
–
FJ826735
EF133613
EF133614
EF133615
EF133616
EF133617
FJ826736
EF133619
EF133618
FJ826737
FJ826738
FJ826739
FJ826740
EF133620
EF133621
–
FJ826741
FJ826742
DQ147811
FJ826628
FJ826629
DQ147813
DQ147814
DQ147815
DQ147817
DQ147816
FJ826630
DQ147819
FJ826631
FJ826632
FJ826633
FJ826885
DQ147822
DQ147823
FJ826634
FJ826635
FJ826636
EF133569
EF133570
EF133571
EF133572
EF133573
FJ826688
EF133622
EF133623
EF133624
EF133625
EF133626
FJ826743
DQ147825
DQ147826
DQ147827
DQ147828
DQ147829
FJ826637
Fischer NE7 (NEU)
Ray Morgan s.n. (LV)
Yuan CN2k1-84 (NEU)
Yuan CN2k-09 (NEU)
Yuan CN2k-30 (NEU)
Knox 3653 (LV)
Song s.n. (NEU)
89.2005 (UC)
EF133574
EF133575
EF133576
EF133578
FJ826689
EF133579
EF133580
EF133627
EF133628
EF133629
EF133630
–
FJ826744
EF133632
EF133633
DQ147830
DQ147832
DQ147833
DQ147835
DQ147837
FJ826638
DQ147839
DQ147840
Knox 3627 (LV)
Knox 3262 (LV)
Jansen 5505 (WAG)
Song S007 (NEU)
Yuan CN2k-41 (NEU)
Menn. and Bar.-Kui. 284
(U)
FB/S3925 (BR)
19860179 (E)
FJ826690
–
–
EF133581
–
FJ826691
FJ826745
FJ826746
–
EF133634
EF133635
–
FJ826639
FJ826640
FJ826641
DQ147841
DQ147842
DQ147843
EF133582
–
EF133636
FJ826747
DQ147845
DQ147846
Yuan CN2k-79 (NEU)
Fischer EF8 (NEU)
Janssens SJ002 (LV)
EF133583
FJ826692
EF133584
EF133637
FJ826748
EF133638
DQ147847
FJ826642
DQ147848
Kew 2398
Knox 3558 (LV)
Janssens SJ006 (LV)
Schwerdtfeger 9492a (B)
Fischer NE10 (NEU)
Knox 2633 (LV)
Knox 4355 (LV)
Knox 2846 (LV)
Dessein SD719 (BR)
FJ826693
FJ826694
EF133586
EF133585
FJ826695
FJ826696
FJ826697
FJ826698
FJ826699
FJ826749
FJ826750
EF133640
EF133639
FJ826751
FJ826752
FJ826753
FJ826754
FJ826755
FJ826643
FJ826644
DQ147850
DQ147849
FJ826645
FJ826646
FJ826647
FJ826648
China, Sichuan
Madagascar
Tanzania
Gabon
Cameroon
Ghana
Hao 426 (NEU)
JBE2 (NEU)
Knox 3571 (LV)
Ngok Banak 2022 (WAG)
J.J.F.E. de Wilde 8638 (BR)
Jongkind 1926 (WAG)
EF133587
FJ826700
FJ826701
FJ826702
FJ826703
FJ826704
EF133641
–
FJ826756
FJ826757
FJ826758
FJ826759
DQ147851
FJ826649
FJ826650
FJ826651
FJ826653
FJ826652
Tanzania
Tanzania
Thailand, Qingmai
Knox 3575 (LV)
Knox 3587 (LV)
Chassot 99–238 (NEU)
FJ826705
FJ826706
–
FJ826760
FJ826761
EF133642
FJ826654
FJ826655
DQ147853
China, Yunnan
China, Yunnan
South Indian origin, cult. by Ray Morgan, UK
African origin, cult. Nat. Bot. Gard. Meise
China, Yunnan
Tanzania
African origin, cult. Bot. Gard. Koblenz Univ.
South Indian origin, cult. by Ray Morgan, UK
China, Yunnan
China, Fujian
China, Yunnan
Tanzania
Vietnam, Anam
Himalayan origin, cult. Univ. California Bot.
Gard. Berkeley
Tanzania
Tanzania
Ethiopia
China, Sichuan
China, Guangxi
Kenya
South Indian origin, cult. Nat. Bot. Gard. Meise
South-African origin, cult. Roy. Bot. Gard.
Edinburgh
China, Yunnan
Madagascar
Belgium, Leuven (introduced; Himalayan
origin)
Seychelles
Tanzania
New Guinean origin
West African origin, cult. Bot. Gard. Berlin
Sri Lanka origin, cult. Bot. Gard. Koblenz Univ.
Kenya
South Africa
Kenya
Zambia
809
S.B. Janssens et al. / Molecular Phylogenetics and Evolution 52 (2009) 806–824
Table 1 (continued)
Taxon
Origin
Voucher
Accession
No.
ImpDEF1
Accession
No.
ImpDEF2
Accession
No.
atpB-rbcL
Impatiens kilimanjari Oliv.
Impatiens latifolia L.
Impatiens lukwangulensis Grey-Wilson
Impatiens mackeyana Hook.f. ssp. claeri (N.Hallé)
Grey-Wilson
Impatiens mackeyana Hook.f. ssp. zenkeri (Warb.)
Grey-Wilson
Impatiens macroptera Hook.f.
Impatiens mannii Hook.f.
Tanzania
South Indian origin, cult. by Ray Morgan, UK
Tanzania
Gabon origin, cult. Nat Bot. Gard. Koblenz Univ.
Knox 3652 (LV)
Ray Morgan s.n. (LV)
Knox 3570 (LV)
Fischer EF5 (NEU)
FJ826707
EF133588
EU723716
FJ826708
EF133643
EU723723
FJ826762
FJ826656
DQ147854
EU723709
FJ826657
African origin, cult. Bot. Gard. Koblenz Univ.
Fischer EF21 (NEU)
FJ826709
–
DQ147857
Equatorial Guinea
East Congo
FJ826710
FJ826711
FJ826763
FJ826764
FJ826658
FJ826659
Impatiens mannii Hook.f.
Impatiens mazumbaiensis Grey-Wilson
Impatiens mengtseana Hook.f.
Impatiens meruensis Gilg ssp. cruciata (T.C.E. Fr.)
Grey-Wilson
Impatiens meruensis Gilg ssp. meruensis
Impatiens mildbraedii Gilg telekii (T. C.E.Fr.)
Grey-Wilson
Impatiens monticola Hook.f.
Impatiens nana Engl.
Impatiens napoensis Y.L. Chen
Impatiens niamniamensis Gilg
Impatiens noli-tangere L.
Impatiens omeiana Hook.f.
Cameroon
Tanzania
China, Yunnan
Kenya
de Wilde 12014 (WAG)
Geuven Carvalho 4534
(BR)
Merckx s.n. (LV)
Knox 3643 (LV)
Yuan CN2k1-38 (NEU)
Ray Morgan s.n. (LV)
FJ826712
EU723717
EF133589
FJ826713
FJ826765
EU723724
EF133644
FJ826766
FJ826660
EU723710
DQ147858
FJ826661
Tanzania
Kenya
Knox 3328 (LV)
Knox 3882 (LV)
FJ826714
FJ826715
FJ826767
FJ826768
FJ826662
FJ826663
China, Sichuan
Tanzania
China, Yunnan
African origin, cult. Nat. Bot. Gard. Meise
Korea
China, Sichuan, cult. Univ. California Bot. Gard.
Berkeley
Tanzania
Hao 425 (NEU)
Knox 2589 (LV)
Yuan CN2k1-61 (NEU)
FB/S2590 (BR)
Janssens SJ008
2002.0214 (UC)
EF133591
EU723719
FJ826716
FJ826717
–
EF133592
EF133646
EU723726
FJ826769
FJ826770
–
EF133647
DQ147860
EU723712
DQ147861
DQ147862
FJ82664
DQ147864
Knox 3572 (LV)
FJ826718
–
FJ826665
Tanzania
Gabon
Indonesian origin, cult. by Ray Morgan, UK
Belgium, Leuven (introduced; Asian origin)
Tanzania
Bali, Indonesian origin, cult. by Ray Morgan, UK
Vietnam, Anam
Tanzania
Gabon
African origin, cult. Roy. Bot. Gard. Edinburgh
Tanzania
China, Yunnan
Tanzania
China, Yunnan
Tanzania
Knox 3552 (LV)
Jongkind 5724 (WAG)
Ray Morgan s.n. (LV)
Janssens SJ004 (LV)
Knox 4420 (LV)
Ray Morgan s.n. (LV)
Song s.n. (NEU)
Knox 3553 (LV)
de Wilde 10390 (WAG)
19680124 (E)
Knox 3229 (LV)
Song Y007 (NEU)
Knox 2592 (LV)
Yuan CN2k1-26 (NEU)
Knox 3462 (LV)
EU723720
FJ826719
FJ826720
EF133593
FJ826721
EF133594
EF133595
FJ826722
FJ826723
EF133597
FJ826724
EF133598
FJ826725
EF133599
FJ826726
FJ826772
FJ826771
EF133648
FJ826773
EF133649
EF133650
FJ826774
FJ826775
–
FJ826776
EF133652
FJ826777
EF133653
–
EU723713
FJ826666
FJ826667
DQ147866
FJ826668
DQ147868
DQ147869
FJ826669
FJ826670
DQ147871
FJ826671
DQ147872
FJ826672
DQ147874
FJ826673
Tanzania
China, Yunnan
Cameroon
Himalayan origin, cult. Holden arboretum
Tanzania
China, Yunnan
Tanzania
Zimbabwe
China, Yunnan
African origin, cult. Bot. Gard. Koblenz Univ.
Tanzania
Kenya
Knox 3667 (LV)
Yuan CN2k1-44 (NEU)
Merckx VM125 (LV)
941314 (DBG)
Knox 3551 (LV)
Yuan CN2 k-80 (NEU)
Knox 3318 (LV)
Knox 4324 (LV)
Yuan CN2 k-57 (NEU)
Fischer NE18
Knox 3576 (LV)
Knox 2639 (LV)
EU723721
EF133600
FJ826727
EF133601
FJ826728
–
FJ826729
EU723722
EF133602
–
FJ826730
-
EU723727
EF133654
FJ826778
EF133655
FJ826779
FJ826780
FJ826781
EU723728
EF133656
FJ826782
FJ826783
FJ826784
EU723714
DQ147876
FJ826666
DQ147877
FJ826674
–
FJ826675
EU723715
DQ147882
DQ147883
FJ826676
FJ826677
China, Yunnan
Madagascan origin, cult. Bot. Gard. Univ.
Kopenhagen
Kenya
Tanzania
China, Yunnan
China, Taiwan
African origin, cult. Bot. Gard. Koblenz Univ.
South Indian origin, cult. by Ray Morgan, UK
Tanzania
African origin, cult. Nat. Bot. Gard. Meise
China, Yunnan
Yuan CN2k1-68 (NEU)
Janssens SJ005 (LV)
EF133603
EF133604
EF133657
EF133658
DQ147885
DQ147886
Knox 2776 (LV)
Knox 3554 (LV)
Yuan CN2k2-173 (NEU)
Zhengyu Jiang T1 (NEU)
Fischer NE20 (NEU)
Ray Morgan s.n. (LV)
Knox 4117 (LV)
S3926 (BR)
Yuan CN2k1-55 (NEU)
FJ826731
FJ826732
EF133605
EF133606
EF133608
FJ826733
EF133609
FJ826734
FJ826785
FJ826786
EF133659
EF133660
EF133661
EF133663
–
EF133664
FJ826787
FJ826678
DQ147887
DQ147888
DQ147889
DQ147891
FJ826679
DQ147892
DQ147894
Impatiens pallide-rosea Gilg ssp. lupangaensis
(G.M. Schulze) Grey-Wilson
Impatiens pallide-rosea Gilg ssp. pallide-rosea
Impatiens palpebrata Hook.f.
Impatiens parasitica Bedd.
Impatiens parviflora DC.
Impatiens percordata Grey-Wilson ssp. percordata
Impatiens platypetala Lindl.
Impatiens poilanei Tardieu
Impatiens pseudohamata Grey-Wilson
Impatiens pseudomacroptera Grey-Wilson
Impatiens pseudoviola Gilg.
Impatiens pseudozombensis Grey-Wilson
Impatiens purpurea Hand.-Mazz.
Impatiens raphidothrix Warb.
Impatiens rectangula Hand.-Mazz.
Impatiens rubromaculata Warb. ssp. imagiensis
Grey-Wilson
Impatiens rubromaculata Warb. ssp. rubromaculata
Impatiens rubrostriata Hook.f.
Impatiens sakeriana Hook.f.
Impatiens scabrida DC.
Impatiens serpens Grey-Wilson
Impatiens siculifer Hook.f.
Impatiens sodenii Engl.
Impatiens sylvicola Burtt Davy and Greenway
Impatiens taronensis Hand.-Mazz.
Impatiens teitensis Grey-Wilson
Impatiens thamnoidea G. M. Schulze
Impatiens tinctoria A. Rich ssp. elegantissima
(Gilg) Grey-Wilson
Impatiens trichosepala Y.L. Chen
Impatiens tuberosa H. Perrier
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
Impatiens
tweedieae E. A. Bruce
ulugurensis Warb.
uliginosa Franch.
uniflora Hayata
usambarensis Grey-Wilson
viscida Wight
volkensii Warb.
walleriana Hook.f.
yingjiangensis S.Akiyama and H.Ohba
of the crown group of Asclepiadoideae (Apocynaceae) constrained
to 38 mya on the basis of the earliest fossil of this group from the
Oligocene and Miocene (Muller, 1981); (3) a minimum age constraint of 87 mya to the Cornaceae based on the fossil fruits of Hiro-
noia fusiformis that were found in the Coniacian stratigraphic zone
(Takahashi et al., 2002; Anderson et al., 2005); (4) minimum age
constraint of 42 mya for Sambucus (Adoxaceae) based on the finding of fossil endocarps from the late Eocene to Pliocene found in
810
S.B. Janssens et al. / Molecular Phylogenetics and Evolution 52 (2009) 806–824
Europe (Reid and Chandler, 1926); (5) based on reports of fossil
fruits of Fraxinus (Oleaceae), occurring in America from the Eocene
onwards, a minimum age for this genus was constrained to
54.9 mya (Call and Dilcher, 1992); (6) minimum age of Weigela
(Caprifoliaceae) constrained to 24.6 mya based on reports of distinctive winged seeds from the Miocene and Pliocene (LancuckaSrodoniowa, 1967); (7) the crown node of Solanaceae is constrained
to a minimum age of 54.9 mya, corresponding to the earliest Solanaceae fossil known from the Eocene (Collinson et al. (1993)); (8)
recent estimation of divergence times based on ndhF and rbcL suggest that Asteraceae have originated in the mid Eocene (Kim et al.,
2005). As a result, the crown node of this clade was constrained at a
minimum age of 47 mya; (9) the lineage towards Valeriana (Valerianaceae) is constrained at a minimum age of 11.2 mya, corresponding with the report of fossil fruits from the late Miocene and
Pliocene in Europe (Bell and Donoghue, 2005); (10) minimum age
of Ixora (Rubiaceae) constrained to 5.3 mya based on the fossil collection of I. casei from the Marshall Islands (Leopold, 1969); (11) in
addition, we assigned a fixed age of 128 mya to crown node of the
asterids, a calibration point estimated by Bremer et al. (2004). In order to determine the impact of the different fossil calibration points
on the overall dating estimation, we used the method of ‘fossil
cross-validation’ developed by Near and Sanderson (2004). Their
approach evaluates the difference between the fossil ages and the
estimated molecular ages by re-evaluating the dating analysis with
only a single calibration point. By calculating the summed square
value (SS) between fossil constraint and molecular estimate, we noticed that calibration points 1 and 2 exhibited the highest scores
(Fig. 1). Removal of these calibration points from our dataset resulted in a significant decline of the average squared deviation (s),
whereas subsequent removal of other fossil constraints did not alter s (Fig. 2). Consequently, we removed fossil constraints 1 and 2,
as they were determined as most inconsistent calibration points
for the asterid dataset.
Confidence intervals for each node were obtained as described
in the r8s manual. PHYLIP (Felsenstein, 1995) was used to generate
100 bootstrapped datasets of the original dataset. For each bootstrapped data matrix, we generated a phylogram with the RAxML
procedure. Using the PL method in r8s 1.70, we recalculated divergence times for each of the phylograms using the same calibration
points as above. These results were then used to estimate standard
deviations for each node.
The obtained age estimate for the split between I. omeiana and
the remainder of the genus was then applied as fixed calibration
point to calculate the divergence events within Impatiens. The
boundaries of the confidence interval for this node were taken into
account in the secondary calibration approach of Impatiens. Hence
we analyzed the Impatiens dataset for an age of 22.5, 22.5 + 5.6 and
22.5 5.6 million years ago. In addition to these indirectly obtained age estimates, we used the following biogeographic age
estimates for cross-validation: The age of Mt Kilimanjaro, a recent
volcanic mountain dated at 1.1 mya harbouring the endemic species Impatiens digitata ssp. digitata and Impatiens kilimanjari (Knox
and Palmer, 1998).
To infer age estimates for the most diversified lineages in Impatiens, we used the combined chloroplast and nuclear data matrix. A
likelihood ratio test showed that a clocklike evolution for the ImpDEF1/ImpDEF2 dataset also had to be rejected. Consequently, we
calculated the divergence times on the combined data using the
penalized likelihood method (Sanderson, 2002) onto the rooted
ML phylogram.
2.4. Ancestral area distribution
In order to reconstruct biogeographic ancestral areas, we used
the likelihood reconstruction method (Pagel, 1999; Schluter
et al., 1997) implemented in Mesquite v1.12 (Maddison and Maddison, 2008). The geographic distribution of Impatiens has been
coded using major geographic units: Africa, Madagascar, Southwest China, the Himalayas, India and Sri Lanka, North America,
Southeast Asia. A Markov k-state one-parameter model (Mk1),
assuming a single rate of transition between two character states
was applied for ML ancestral area reconstruction. In order to determine the best estimate of the reconstructed characters state at
each node, a likelihood ratio test was used with a likelihood decision limit of 2.0. With log-likelihoods differing by 2 or more, the
lowest negative log-likelihood was regarded as best value, whereas
nodes with log-likelihoods differences lower than two were assumed to be ambiguous.
2.5. Speciation rates and timing of diversification in the genus
Impatiens
One aspect of this study is to examine whether the timing of
speciation in Impatiens is correlated with climate changes during
the Pliocene–Pleistocene. By generating a lineages-through-time
(LTT) plot for the PL chronogram derived from the combined Impatiens dataset, we obtain a schematic visualization of the net diversification rate of the genus Impatiens. The outline of the LTT plot is
used to evaluate whether changes in diversification rate have occurred during the history of Impatiens. Only one individual per species was included in the LTT plot as the inclusion of individuals
representing additional subspecies would inflate the speciation
rate. Present sampling only contains about 11% of all known Impatiens species (113 out of 1000 species). As a result, we tested the
consequence of an incomplete taxon sampling on the profile of
the empirical LTT plot by generating simulated phylogenies with
the software program Phylogen 1.1 (Rambaut, 2002). The following
simulation was executed. A total of 1000 replicate phylogenies
were generated, containing 1000 extant taxa of which 884 were
randomly pruned from each tree. Each simulation was generated
under the assumption of a constant birth–death rate (b/d = 0).
Branch lengths of the resulting phylograms were rescaled with
TreeEdit 1.0 (Rambaut and Charleston, 2002) to 22.5 million years,
the estimated crown node age of Impatiens. We subsequently evaluated the empirical LTT curve by comparison with the mean LTT
simulation and its 95% confidence interval.
In order to assess a quantitative method that could illustrate an
overall change in diversification rate within the evolutionary history of Impatiens, the rate of speciation was calculated for each
lineage ([ln(N) ln(N0)/T]; Baldwin and Sanderson, 1998). Additionally, we used the constant rate (CR) test of Pybus and Harvey
(2000) to investigate a possible increase in overall diversification
rate within Impatiens. The CR method as implemented in Genie
3.0 (Pybus and Harvey, 2000) follows an approach of c-statistics
in which the relative position of the nodes in the empirical phylogeny is compared to the position of those nodes that are assumed to
be under a model of constant diversification. Positive c-values
(c > 0) indicate an apparent acceleration of the overall diversification rate, assuming that nodes are closer to the tips than expected
under a pure-birth–death model (c = 0). On the contrary, negative
c-values (c < 0) display a significant decrease in diversification
rate. The null hypothesis of constant birth–death rate can be rejected when c < 1.645 (one-tailed test). Because the c-statistics
is biased by extinction and incomplete taxon sampling, we simulated pure-birth topologies (d/b = 0) using Phylogen 1.1 (Rambaut,
2002). Analyses were carried out under the assumption that our
Impatiens sampling (n = 116) only represents a small fraction (f)
of the real diversity of the genus. We simulated 1000 pure-birth
topologies with n/f tips, with the f-scores varying from 0.06 to 1.
The resulting trees were then pruned to 117 taxa and for each of
the pruned trees the c-statistics was computed using Genie 3.0
811
S.B. Janssens et al. / Molecular Phylogenetics and Evolution 52 (2009) 806–824
Table 2
GenBank accessions of atpB-rbcL spacer sequences of the asterid used in the large age estimation analysis.
Species
Accession No.
Family
Order
Kirengeshoma palmata
Cornus mas
Actinidia chinensis
Hydrocera triflora
Impatiens
Diospyros kaki
Euclea pseudebenus
Lissocarpa guianensis
Acrothamnus hookeri
Acrothamnus maccraei
Acrotriche ramiflora
Andersonia sprengelioides
Andromeda polifolia
Androstoma empetrifolia
Archeria comberi
Archeria serpyllifolia
Astroloma xerophyllum
Brachyloma daphnoides
Brachyloma preissii
Bruckenthalia spicufolia
Budawangia gnidioides
Chamaedaphne calyculata
Coleanthera myrtoides
Conostephium preissii
Cosmelia rubra
Croninia kingiana
Cyathodes platystoma
Cyathodes pumila
Cyathopsis floribunda
Daboecia cabtabrica
Diplycosia cinnamomifolia
Dracophyllum longifolium
Epacris obtusifolia
Erica hispidula
Gaultheria tomentosa
Harrimanella hypnoides
Kalmia latifolia
Leptecophylla tameiameiae
Leucopogon milliganii
Leucothoe fontanesiana
Lissanthe synandra
Lyonia lucida
Lysine ciliatum
Melichrus urceolatus
Melichrus urceolatus
Monotoca elliptica
Oligarrhena micrantha
Pentachondra involucrata
Pernettya tasmanica
Planocarpa petiolaris
Prionotes cerinthoides
Rhododendron simsii
Richea pandanifolia
Rupicola sprengelioides
Sphenotoma dracophylloides
Styphelia tubiflora
Trochocarpa laurina
Woollsia pungens
Zenobia pulverulenta
Corytophora alta
Eschweilera simiorum
Marcgravia maguirei
Marcgravia umbellata
Norantea guianensis
Souroubea sp.
Primula cuneifolia
Chromolucuma rubriflora
Chrysophyllum imperiale
Eberhardtia aurata
Elaeolum glabrescens
Changiostyrax dolichocarpa
Pterostryrax psilophyllus
Sinojackia sarcocarpa
Pelliciera rhizophorea
Aucuba japonica
EF437413
X83988
FJ866479
DQ147895
DQ147810–DQ147894
X91004
FJ866480
AF421094
AY971370
AY971371
AY636036
AF155843
AF366584
AY372540
AF155840
AY971372
AY372554
AF155859
AY372555
AY520757
AF155852
AF366616
AY372556
AY372557
AF155842
AF208750
AY372560
AY372558
AY636040
AY520758
AF366587
AF155845
AY636042
AY520767
AF366611
AF155833
AB247970
AY372569
AY372575
AF366612
AY636038
AF155836
AF155848
AY372595
AY372595
AY005085
AF155854
AY005087
AF155835
AY372594
AF155838
AM296078
AF155844
AF155851
AF155846
AY372591
AY005092
AF155847
AF366615
AF076769
X91001
DQ147896
DQ147897
DQ147898
DQ147900
AB003575
EF558591
EF558592
EF558594
EF558593
DQ317984
DQ317980
DQ317979
DQ147899
AB087779
Hydrangeaceae
Cornaceae
Actinidiaceae
Balsaminaceae
Balsaminaceae
Ebenaceae
Ebenaceae
Ebenaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Ericaceae
Lecythidaceae
Lecythidaceae
Marcgraviaceae
Marcgraviaceae
Marcgraviaceae
Marcgraviaceae
Primulaceae
Sapotaceae
Sapotaceae
Sapotaceae
Sapotaceae
Styracaceae
Styracaceae
Styracaceae
Tetrameristaceae s.l.
Garryaceae
Cornales
Cornales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Ericales
Garryales
Superorder
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Basal Asterids
Euasterids 1
(continued on next page)
812
S.B. Janssens et al. / Molecular Phylogenetics and Evolution 52 (2009) 806–824
Table 2 (continued)
Species
Accession No.
Family
Order
Superorder
Garrya ovata
Alstonia scholaris
Dischidia astephana
Gunessia pepo
Hoya kentiana
Madangia inflata
Marsdenia carvalhoi
Micholitzia obcordata
Gentiana oreodoxa
Mitrasacme pygmaea
Mitreola petiolata
Adina rubella
Alberta minor
Argostemma hookeri
Asperula cynanchica
Augusta longifolia
Aulacocalyx jasminiflora
Belonophora coriacea
Bertiera dewevrei
Cephalanthe occidentalis
Chiococca alba
Chomelia brasialiana
Cinchona pubescens
Coffea congensis
Colletoecema dewevrei
Commitheca liebrechtis
Coptosapelta diffusa
Coryanthe mayumbensis
Cosmocalyx spectabilis
Coussarea hirticalyx
Crossopteryx febrifuga
Cruciata glabra
Cruckshansia hymencodon
Cubanola domingensis
Dictyandra arborescens
Didymosalpinx lanciloba
Diplospora dubia
Duperrea pavettifolia
Ecpoma hierniana
Eosanthe cubensis
Erithalis fruticosa
Euclinia longiflora
Exostema lineatum
Faramea porophylla
Fernelia buxifolia
Gaertnera sp.
Galium scabrum
Gardenia imperialis
Guettarda platypoda
Hayataella michelloides
Hedyotis littoralis
Heinsia crinita
Hilllia triflora
Hoffmannia refulgens
Hydnophytum formicarum
Isertia rosea
Ixora coccinea
Ixora finlaysonia
Keetia multiflora
Kellogia galioides
Ladenbergia sp.
Leptactina leopoldi-secundi
Luculia pinceana
Manettia cordifolia
Mitragyne inermis
Morelia senegalensis
Morinda lucida
Mycetia malayana
Myrmecoda platytyrea
Myrmephytum beccarii
Otiophora scabra
Paratriaina xerophila
Pausinystalia johimbe
Pavetta ternifolia
Pentas lanceolata
Phialanthus stillans
AB087788
DQ359161
DQ334576
DQ334570
DQ334564
DQ334583
DQ334563
DQ334601
DQ398621
DQ131694
DQ131696
DQ131698
DQ131699
AJ234032
X81689
DQ131703
DQ131704
DQ131706
DQ131707
DQ131710
DQ131711
DQ131712
AJ233990
AM412393
DQ131713
AJ233999
AJ233987
DQ131715
DQ131716
DQ131717
DQ131719
X81097
AJ234004
DQ131720
DQ131723
DQ131724
DQ131725
DQ131726
DQ131727
DQ131729
DQ131730
DQ131731
DQ131732
AJ234008
DQ131736
AJ234012
X76462
DQ131737
DQ131739
AB247242
AJ234027
DQ131740
AJ233993
X81684
X76480
DQ131743
AM412400
DQ131744
DQ131745
AY570768
DQ131747
DQ131748
DQ131749
AJ234023
DQ131751
DQ131752
DQ131753
AJ234033
AB044151
AB044152
DQ131756
DQ131759
DQ131760
DQ131761
AB247149
DQ131762
Garryaceae
Apocynaceae
Apocynaceae
Apocynaceae
Apocynaceae
Apocynaceae
Apocynaceae
Apocynaceae
Gentianaceae
Loganiaceae
Loganiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Garryales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
813
S.B. Janssens et al. / Molecular Phylogenetics and Evolution 52 (2009) 806–824
Table 2 (continued)
Species
Accession No.
Family
Order
Plocama pendula
Pomax umbellata
Portlandia grandiflora
Pouchetia africana
Pseudosabicea floribunda
Psilanthus leroyi
Psychotria undata
Putoria calabrica
Remijia macronemia
Rothmannia sp.
Rubia tinctorum
Rudgea bremekampiana
Rutidea schlechteri
Sabicea diversifolia
Sacosperma paniculatum
Salzmannia axillaris
Sarcocephalus latifolius
Schumanniophyton magnificum
Serissa foetida
Spermacoce assurgens
Stelechante cauliflora
Stipularia elliptica
Tricalysia elliotii
Trichostachys perfoliatum
Urophyllum glabrum
Valantia muralis
Vangueria infausta
Vangueria infausta
Wendlandia sp.
Achimenes admirabilis
Aeschynanthus longiflorus
Alsobia dianthiflora
Besleria melancholica
Chirita longgangensis
Chirita spadiciformis
Chrysothemis pulchella
Cobanantha calochlamys
Codonantha serrulata
Columnea sanguinea
Cyrtandra cupulata
Didissandra frutescens
Didymocarpus citrinus
Didymocarpus cordatus
Emarhendia bettiana
Episcia cupreata
Epithema taiwanese
Fieldia australis
Haberlea rhodopensis
Henckelia albomarginata
Jovellana punctata
Koellikeria erinoides
Kohleria eriantha
Lenbrassia australiana
Loxonia hirsutum
Mitraria coccinea
Monophyllae horsfieldii
Nautilocalyx melittifolius
Nenatanthus villosus
Paliavana sericiflora
Paraboea capitata
Petrocosmea nervosa
Primulina tabacum
Ramonda myconi
Rhabdothamnops sinensis
Rhynchoglossum obliquum
Rhytiophyllum tomentosumbesleria
Saintpaulia velutina
Sarmiaeta scandens
Sinningia bulbosa
Stauranthera grandiflora
Streptocarpus holstii
Streptocarpus rexii
Titanotrichum oldhamii
Trisepalum speciosum
Vanhouttea gardneri
Whytockia sasakii
AJ234035
DQ131767
DQ131768
DQ131770
DQ131771
AM412399
DQ131774
X81672
DQ131775
DQ131776
X76474
DQ131778
DQ131779
DQ131781
DQ131782
DQ131784
DQ131785
DQ131786
AJ234034
X81679
DQ131789
DQ131790
DQ131791
AF446995
DQ131793
X76473
DQ131794
DQ131794
DQ131796
AJ439982
AJ490920
AJ490924
AJ490923
AJ490903
AJ490904
AY423115
AJ490926
AJ439981
AJ490927
AJ490886
U91313
AJ490906
AJ490907
AJ490908
AJ490928
AY423117
AY423112
AJ490909
AJ490910
AY423109
AJ439983
AY423114
AJ490921
AJ490891
AY423113
U91315
AJ439984
AJ439980
AJ439963
AJ490911
AJ490912
AJ490913
AJ490914
AJ490915
AJ490898
AJ490930
AJ490916
AJ490922
AJ439910
AJ490900
AJ490917
AJ490918
AY423111
AJ490919
AJ439974
AY423116
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Rubiaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gesneriaceae
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Gentianales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Superorder
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
Euasterids 1
(continued on next page)
814
S.B. Janssens et al. / Molecular Phylogenetics and Evolution 52 (2009) 806–824
Table 2 (continued)
Species
Accession No.
Family
Order
Superorder
Globularia salicina
Phlomis crinita
Utricularia gibba
Fraxinus excelsior
Plantago media
Antirrhinum majus
Calceolaria arachnoidea
Digitalis purpurea
Gratiola aurea
Paulownia tomentosa
Scrophularia canina
Tetranema mexicanum
Verbascum speciosum
Veronica incana
Aureliana fasciculata
Capsicum baccatum
Datura stramonium
Jaltomata auriculata
Lycianthe lenta
Lycopersicon pimpinellifolium
Nicandra physalodes
Schizanthus x wisetonensis
Solanum shanesii
Tubocapsicum anomalum
Wythania coagens
Chaerophyllum nodosum
Myrrhis odorata
Oreomyrrhis azorellaceae
Osmorhiza longistylus
Aralia cachemirica
Brassaiopsis shweliensis
Eleutherococcus trifoliatus
Fatsia japonica
Harmsiopanax aculeata
Hedera maderensis
Heteropanax fragrans
Macropanax dispermus
Merilliopanax cordifolius
Panax assamicus
Polyscias javanica
Schleffera hypoleucoides
Tupidanthus calyptratus
Helwingia japonica
Ilex discolor
Achillea millefolium
Bellis perennis
Ligularia franchetiana
Apetahia margaretae
Brighamia insignis
Burmeistera crispiloba
Centropon gutierrezii
Clermontia fauriei
Cyanea pilosa
Delissea subcordata
Isotoma axillaris
Lobelia hypoleuca
Nymphoides cordata
Sclerotheca jayorum
Trematolobelia macrostachys
Adoxa moschatellina
Sambucus racemosa
Sinadoxa corydalifolia
Tetradoxa omeiensis
Viburnum acerifolium
Leycesteria formosa
Lonicera sempervirens
Symphoricarpos orbiculatus
Triosteum perfoliatum
Triplostegia glandulifera
Diervilla sessilifolia
Weigela hortensis
Dipsacus mitis
Knautia macedonica
Scabiosa columbaria
Heptacodium miconioides
Abelia x grandiflora
AY818898
AY792758
EF529719
AY911655
AY818903
AJ490883
AY423108
AY818897
EF529725
AY423104
AY423105
AJ490884
AJ490885
AY818908
AF397083
AF397101
AF397076
AF397081
AF397093
AF397079
AJ490882
AY423103
AF397088
AF397082
AF397084
DQ829733
DQ829738
DQ829705
DQ829737
AY753224
AY753233
AY753235
AY163545
AY753251
AY163533
AY753236
AY753241
AY753243
AY753245
AY753252
AY753247
AY753250
X94941
AF471629
EU129084
X91000
AB375447
DQ285286
DQ285256
DQ285281
DQ285282
DQ285259
DQ285291
DQ285264
DQ285283
DQ285266
EF529716
DQ285273
DQ285271
AF446990
AF446988
AF446989
AF446991
AF446987
EU265521
EU265563
AF446994
AF446995
AF447009
AF446997
AF446998
AF447007
AY362501
AF447008
AF446996
AF446999
Globulariaceae
Lamiaceae
Lentibulariaceae
Oleaceae
Plantaginaceae
Scrophulariaceae
Scrophulariaceae
Scrophulariaceae
Scrophulariaceae
Scrophulariaceae
Scrophulariaceae
Scrophulariaceae
Scrophulariaceae
Scrophulariaceae
Solanaceae
Solanaceae
Solanaceae
Solanaceae
Solanaceae
Solanaceae
Solanaceae
Solanaceae
Solanaceae
Solanaceae
Solanaceae
Apiaceae
Apiaceae
Apiaceae
Apiaceae
Araliaceae
Araliaceae
Araliaceae
Araliaceae
Araliaceae
Araliaceae
Araliaceae
Araliaceae
Araliaceae
Araliaceae
Araliaceae
Araliaceae
Araliaceae
Helwingiaceae
Aquifoliaceae
Asteraceae
Asteraceae
Asteraceae
Campanulaceae
Campanulaceae
Campanulaceae
Campanulaceae
Campanulaceae
Campanulaceae
Campanulaceae
Campanulaceae
Campanulaceae
Menyanthaceae
Campanulaceae
Campanulaceae
Adoxaceae
Adoxaceae
Adoxaceae
Adoxaceae
Adoxaceae
Caprifoliaceae
Caprifoliaceae
Caprifoliaceae
Caprifoliaceae
Caprifoliaceae
Diervillaceae
Unplaced
Dipsacaceae
Dipsacaceae
Dipsacaceae
Unplaced
Linnaeaceae
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Lamiales
Solanales
Solanales
Solanales
Solanales
Solanales
Solanales
Solanales
Solanales
Solanales
Solanales
Solanales
Apiales
Apiales
Apiales
Apiales
Apiales
Apiales
Apiales
Apiales
Apiales
Apiales
Apiales
Apiales
Apiales
Apiales
Apiales
Apiales
Apiales
Aquifoliales
Aquifoliales
Asterales
Asterales
Asterales
Asterales
Asterales
Asterales
Asterales
Asterales
Asterales
Asterales
Asterales
Asterales
Asterales
Asterales
Asterales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
815
S.B. Janssens et al. / Molecular Phylogenetics and Evolution 52 (2009) 806–824
Table 2 (continued)
Species
Accession No.
Family
Order
Superorder
Dipelta yunnanensis
Kolkwitzia amabilis
Linnaea borealis
Zabelia biflora
Acanthocalyx albus
Cryptothladia chinensis
Morina longifolia
Centranthus ruber
Fedia graciliflora
Nardostachys jatamansii
Plectritis macrocera
Valeriana bractescens
Valeriana celtica
Valerianella locusta
Castanea crenata
Vatairea fusca
AF447000
AF447002
AF447001
EU265530
AF447003
AF447004
AF447005
AF448572
AF448575
AF447010
AF447015
AF448580
AY362516
AF447014
AB124938
EF466253
Linnaeaceae
Linnaeaceae
Linnaeaceae
Linnaeaceae
Morinaceae
Morinaceae
Morinaceae
Valerianaceae
Valerianaceae
Valerianaceae
Valerianaceae
Valerianaceae
Valerianaceae
Valerianaceae
Rosids
Rosids
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Dipsacales
Rosids
Rosids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Euasterids
Outgroup
Outgroup
3500
3000
2500
SS
2000
1500
1000
500
0
1
2
3
4
5
6
7
8
9
10
Fossil calibration
Fig. 1. Histogram of the summed square values (SS) of the deviations between
estimated molecular and fossil ages for each calibration point.
45
40
35
30
s
25
20
15
10
5
0
1
2
3
4
6
5
7
Fossils removed
8
9
10
Fig. 2. Plot illustrating the effect of the removal of fossil calibration points on the
average squared deviation (s).
(Pybus and Rambaut, 2002). In addition, we calculated the 95%
confidence interval and the mean c-value for each of the simulated
datasets.
3. Results
3.1. Sequence characteristics and phylogenetic results
Sequence characteristics for all datasets are summarized in
Table 3. Ambiguously aligned nucleotides were removed from both
chloroplast and nuclear data matrices. Although some loci have not
been amplified from a few species, their absence probably did not
2
2
2
2
2
2
2
2
2
2
2
2
2
2
alter the analyses when coded as missing data in the combined
data matrix. A recent study demonstrated that highly incomplete
taxa can be accurately placed in phylogenies as long as many characters were sampled throughout (Wiens, 2005).
The ML analysis of the enlarged chloroplast Impatiens-asterids
dataset yielded a well-supported phylogenetic hypothesis in which
the different orders are well defined and the relationship between
the orders largely corroborates the study of Bremer et al. (2002).
Nevertheless, we observed some discrepancies between their phylogeny and the present study, yet all of which are weakly supported in both studies. Whereas our results reveal a sistergroup
relationship between Solanales and Gentianales (ML-BS: 56), and
Asterales and Dipsacales (ML-BS: 73), the study of Bremer et al.
(2002) shows a closer affinity between Solanales and Lamiales
(maximum parsimony-BS: <50) on the one hand, and an unresolved polytomy with Dipsacales, Asterales and Apiales on the
other. The phylogeny of Bremer et al. (2004) however, which is
based on the molecular results of Bremer et al. (2002) indicates a
sistergroup relationship between Dipsacales and Asterales.
Within the order Ericales, the balsaminoid clade is sister to the
other Ericalean families (ML-BS: 100; e.g. Bremer et al., 2002;
Schönenberger et al., 2005). In contrast to several molecular studies on Ericales and asterids, in which the mutual relationships between Balsaminaceae, Tetrameristaceae s.l. and Marcgraviaceae
remained unresolved (APG, 2003; Savolainen et al., 2000; Schönenberger et al., 2005), the present study reveals a well supported sistergroup relationship between Tetrameristaceae s.l. (Pelliciera) and
the Balsaminaceae–Marcgraviaceae clade (ML-BS: 98). Although
many of the early diversified nodes in Impatiens are well resolved,
the most recent nodes within the genus show limited resolution
(Fig. 3). However, the enlarged asterid dataset was put together
to resolve the basal relationships within the balsaminoid clade,
which are subsequently used in the time divergence estimation
procedure.
ML analysis of the nuclear ImpDEF1/ImpDEF2 data matrix of
Impatiens yielded highly congruent trees, supported by moderate
bootstrap support (data not shown). Topologies based on ImpDEF1/ImpDEF2 are slightly better resolved than those based on
plastid atpB-rbcL sequences. The separately analyzed plastid and
nuclear datasets of Impatiens corroborated the major clades recognized by Janssens et al. (2006) with no incongruent clades found
between the two major data sets. Furthermore, the partition
homogeneity test was not significant (P > 0.05), showing that chloroplast and nuclear partitions of the combined data were not in
conflict. For the combined data matrix, the ML search yielded a
well-resolved topology with strong support for most of the clades
(Fig. 4). In comparison to the separate ImpDEF1/ImpDEF2 and atpBrbcL analyses, the combined phylogeny is better resolved and support values are generally higher.
816
S.B. Janssens et al. / Molecular Phylogenetics and Evolution 52 (2009) 806–824
Table 3
Sequence characteristics of the nuclear and chloroplast loci and the combined matrix.
Analyzed markers
Analyzed characters
Variable characters
Parsimony informative characters
Asterid dataset
Impatiens dataset
atpB-rbcL
ImpDEF1/ImpDEF2
atpB-rbcL
Combined
3822
908
687
5637
804
468
989
248
133
6626
1052
601
100
3
Cornales
Impatiens
Balsaminaceae
100
Ericales
100
Balsaminoids
98
100
100
11
Marcgraviaceae
Tetrameristaceae s.l.
75
1
Lamiales
89
5
97
7
100
Solanales
100
10
56
Gentianales
96
100
99
2
100
Garryales
Aquifoliales
100
95
73
Asterales
8
93
95
Apiales
9
96
Dipsacales
6
68
140
120
100
4
80
60
Rosids
40
20
0
Fig. 3. ML phylogram of the asterid atpB-rbcL dataset. The considered calibration points (1–11) are indicated on the tree. ML bootstrap support values for the most basal
nodes until the nodes at order level are highlighted on the branches. Support values at lower taxonomical level are not shown.
S.B. Janssens et al. / Molecular Phylogenetics and Evolution 52 (2009) 806–824
3.2. Divergence time estimates
Using the ‘fossil cross-validation’ method of Near and Sanderson (2004) we identified the Ebenaceae fossil A. cryptostoma and
817
the earliest known Asclepioid fossil as most inconsistent calibration points in our analysis. The node to which A. cryptostoma was
assigned has been dated 11 million years younger than assumed
by fossil data, whereas the crown node of the Asclepioideae was
Fig. 4. ML phylogram based on the combined chloroplast and nuclear dataset of Impatiens. Asterisks indicate the presence of collapsed branches in strict consensus tree.
Numbers on the branch represent ML bootstrap support values. Clades M (Madagascar), SEA1 (Southeast Asia), SEA2 (Southeast Asia), SI1 (South India), SI2 (South India), NA
(NorthAmerica), A1 (Africa), A2 (Africa) and A3 (Africa) are indicated with an arrow. Blue bars indicate age intervals.
818
S.B. Janssens et al. / Molecular Phylogenetics and Evolution 52 (2009) 806–824
Table 4
Estimated ages for the crown and stem groups of the asterid orders.
Order
Penalized likelihood age estimates (Mya)
Bremer et al. (2004) age estimates (Mya)
Stem
Crown
Stem
Crown
Cornales
Ericales
Garryales
Lamiales
Gentianales
Solanales
Aquifoliales
Apiales
Dipsacales
Asterales
128*
123 ± 10.5
112 ± 9.3
104 ± 8.2
101 ± 7.9
101 ± 11.8
113 ± 9.8
104 ± 11.2
105 ± 14.2
104 ± 12.1
104 ± 13.1
117 ± 9.2
20 ± 8.6
95 ± 11.9
79 ± 10.2
58 ± 9.1**
62 ± 11.9***
87 ± 14.1
99 ± 15.4
94 ± 11.2
128*
127
114
106
108
106
121
113
111
112
112
114
–
97
78
100
113
84
101
93
*
**
The crown node of the asterid was fixed at 128 Mya.
The Convolvulaceae were not included in our dating analysis, which explains the young age estimate for the crown group of the Solanales.
The Stemonuraceae and Cardiopteridaceae were not included in our dating analysis, which explains the young age estimate for the crown group of the Aquifoliales.
***
estimated approximately 15 million years younger than suggested by fossil data. However, when including both fossil calibration points in the overall dating analysis, the dating
estimates of the Impatiens nodes were hardly altered. Nevertheless, both calibration points were removed from the final dating
analysis.
Applying the divergence times of the selected asterid fossils
and the previously estimated age of the asterid crown group by
Bremer et al. (2004), we dated the split between Marcgraviaceae
and Balsaminaceae at 48.2 mya (SD = ±9.3 mya). The divergence
between Impatiens and Hydrocera is estimated at 30.7 mya
(SD = ±8.6 mya), whereas the crown group of Impatiens is estimated at 22.5 mya (SD = ±5.6 mya). The age estimations of the
asterid orders with their standard deviations are listed in Table
4. In order to infer the divergence times for the remaining lineages in Impatiens, we used the estimated crown age of Impatiens
(22.5 my) as secondary calibration points for the ML phylogram
of the combined nuclear-plastid dataset (Fig. 4). In addition, geological evidence was used to cross-validate the asterid-based
divergence times.
3.3. Ancestral area distribution
Maximum likelihood-based ancestral area reconstruction illustrates that the vast majority of the current diversification of Impatiens has originated in Asia (Fig. 5 and Table 5). Southwest China
was unambiguously reconstructed as the ancestral area for the
earliest-diverging lineages. Although likelihood ratio values were
in general significant for the Impatiens topology, we noticed that
the root node, which connects Impatiens to its sistergroup Hydrocera, has an insignificant likelihood ratio value. Apparently, this
insignificance occurs rather frequently in likelihood-based ancestral state reconstruction analyses as the amount of uncertainty
related with reconstruction increases in time (Schluter et al.,
1997).
Biogeographic reconstructions indicate that Impatiens dispersed
into South India via two independent colonization events from
Southwest China (clades SI1 and SI2), whereas the African continent was colonized from Southwest China in three independent
dispersal events (clades A1, A2 and A3).
Similarly, our reconstructions show that only one speciation
event from Southwest China accounts for the diversity of Impatiens
in North America (clade NA). Furthermore, the Southeast Asian
(clades SEA1 and SEA2) and Himalayan species originated from
Southwest China. In contrast, the Malagasy clade (clade M) is
derived from a single colonization event from Africa (Fig. 5 and
Table 5).
3.4. Speciation rates and timing of diversification in Impatiens
The slope of the LTT plot obtained from the PL tree for Impatiens
is slightly convex until 4.5 million years ago and then becomes
steeper (Fig. 6). Simulated LTT curves for Impatiens diversity indicate a hypothesized steep slope near the root of the genus that
gradually decreases towards the tips. Interestingly, between
15 million years ago and present-day, the empirical LTT plot is situated outside the 95% significance level of the simulated LTT plots.
In order to statistically test whether this observation truly indicates a late increase in net diversification rate, we evaluated our results with c-statistics (Fig. 7). For our chronogram, we computed a
negative c-value (c = 6.27), indicating that the process of lineage
accumulation in Impatiens did not remain constant in time. Simulations carried out to consider the effect of incomplete sampling
(113 species sampled from 1000 extant species), indicated that under a true value of 0, Impatiens would only contain 207 species
with the 95% interval, ranging between 120 and 427 species to obtain a c-value as extreme as 6.27. As a result, we are able to reject
the hypothesis that the obtained c-value was the result of a poor
sampling density.
Similar results are obtained when the diversification rate was
calculated for every possible node and its accompanying lineage
(Fig. 8). Between 22.5 and 5 mya, computed speciation rates are
estimated between 0.03 and 0.56 species per million years. During
the next 5 my, rates of species diversification rates fluctuate between 0.07 and 3.8 species per million years.
4. Discussion
4.1. Divergence time estimates in asterids
Table 4 lists the stem and crown group estimates for all asterid
orders. Although the obtained divergence estimates generally corroborate the results of Bremer et al. (2004), the majority of the
stem group estimates are slightly younger, whereas most of the
crown group estimates are a little older. The largest differences
in age estimate between our study and the study of Bremer et al.
(2004) were observed for the stem node age of the orders in the
euasterids 2 clade and the crown group age of the orders Solanales,
Cornales and Aquifoliales. These larger dissimilarities are possibly
due to the use of a single gene marker (atpB-rbcL) that was not applied in the multigene dating analysis on the asterids by Bremer
et al. (2004). In addition, we were unable to sample all the earliest-diverging lineages of each order, resulting in much younger
crown node estimates for some of the above-mentioned lineages.
The use of a single gene for generating age estimates also caused
S.B. Janssens et al. / Molecular Phylogenetics and Evolution 52 (2009) 806–824
819
Fig. 5. ML topology illustrating the reconstruction of ancestral area distribution of Impatiens ancestors. Pie charts at each node indicate proportional likelihoods of the most
common recent ancestor for each clade. Clades M (Madagascar), SEA1 (Southeast Asia), SEA2 (Southeast Asia), SI1 (South India), SI2 (South India), NA (North America), A1
(Africa), A2 (Africa) and A3 (Africa) are indicated with an arrow.
the bootstrap confidence intervals to be fairly large, indicating a
certain inaccuracy of the atpB-rbcL dataset. However, the overall
similarity in ordinal age estimates in Bremer et al. (2004) and
the present study suggests that our chronogram presents a
plausible scenario for other nodes among the asterids (e.g.
Impatiens).
Despite the slightly younger age estimate for most of the stem
nodes of the asterid orders, the crown group of the Ericales, to
820
S.B. Janssens et al. / Molecular Phylogenetics and Evolution 52 (2009) 806–824
Table 5
Proportional likelihoods values for reconstruction of ancestral areas. Values in bold indicate best estimates. Node numbers correspond to those shown in Fig. 5.
NODE
Africa
Madagascar
Himalaya
Southwest China
North America
Southeast Asia
South India-Sri Lanka
Inferred ancestral area
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
6.5
5.6
6.1
99.1
99.0
0.8
0.4
0.1
<0.05
0.1
<0.05
1.9
<0.05
98.2
<0.05
0.9
<0.05
<0.05
99.1
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
0.8
99.0
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
1.3
2.5
3.7
3.7
2.6
93.3
94.2
93.7
0.8
<0.05
<0.05
99.1
98.4
0.8
98.5
0.8
95.2
0.8
1.6
<0.05
99.0
98.4
1.7
0.8
99.1
98.6
97.4
96.2
96.2
97.3
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
1.5
98.2
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
0.1
0.1
<0.05
<0.05
<0.05
0.2
1.3
<0.05
<0.05
99.0
2.8
99.1
<0.05
0.9
<0.05
<0.05
<0.05
<0.05
0.8
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
0.1
<0.05
99.1
1.2
<0.05
<0.05
<0.05
<0.05
99.0
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
Southwest China
Southwest China
Southwest China
Africa
Africa
Madagascar
Southwest China
Southwest China
South India-Sri Lanka
Southwest China
Southeast Asia
Southwest China
Southeast Asia
Africa
South India-Sri Lanka
Southwest China
Southwest China
North America
Africa
Southwest China
Southwest China
Southwest China
Southwest China
Southwest China
Southwest China
which Impatiens belongs, was dated three million years older
(117 ± 10.5 my) than calculated by Bremer et al. (2004). In addition, the study of Bremer et al. (2004) estimated the crown group
of the balsaminoid clade at approximately 64 million years, a date
that corresponds with our results (58.9 my). The crown group age
of the balsaminoids is also slightly older than the first known fossil
of Pelliciera rhizophorae from the Late Eocene (54.9 mya; Graham,
1975; Rull, 1999). We decided not to apply this age estimate in
the asterid dating analysis but to use the fossil pollen age of Pelliciera to cross-validate our obtained results. By applying the fossil
age of P. rhizophorae to the node directly leading towards Balsaminaceae and Marcgraviaceae, the estimated age of Impatiens would
have been strongly influenced by this one calibration point,
6
4
3
2
Number of reconstructed
lineages (ln)
5
1
0
20
15
10
5
0
Million years before present
Miocene
Pliocene Pleistocene
whereas all other fossil calibration points would have had no value
in dating Impatiens.
4.2. Origin and evolution of the Balsaminaceae
Only recently, molecular evidence has demonstrated that Balsaminaceae are closely related to Marcgraviaceae, both being part of
the balsaminoid clade in the Ericales (Bremer et al., 2002; Schönenberger et al., 2005). Despite their close affinity, these families have
very different distribution areas. The Balsaminaceae have a worldwide distribution except for Australia and South America, whereas
the Marcgraviaceae are restricted to tropical forests of South and
Central America (Dressler, 2004; Fischer, 2004). It seems plausible
that their most recent common ancestor was present by the start of
the Tertiary (Grey-Wilson, 1980a). Although the split between
Marcgraviaceae and Balsaminaceae is dated in this study in the
Middle Eocene (48.1 mya), the ancestral stock of both families already existed during the Middle Paleocene (58.9 mya). This
Proportion (f) of Impatiens sampled
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
0
-2
-4
-6
Average γ -value=-6.27
-8
-10
-12
-14
Fig. 6. Semilogarithmic lineage-through-time (LTT) plot for the Impatiens clade
sister to I. omeiana are indicated by a black line. LTT plot from simulated
phylogenies illustrating the effect of an incomplete sampling is shown by a grey
line. The thin grey lines indicate the 95% interval while the thick grey line
designates a mean value.
Fig. 7. Relationship between c and the proportion of extant taxa in Impatiens (f).
The thin grey lines indicate the 95% interval while the thick black line appoints the
mean c-value. The shaded area represents the range of f-values consistent with the
observed c at the 95% level.
821
S.B. Janssens et al. / Molecular Phylogenetics and Evolution 52 (2009) 806–824
3.5
1
3
2
2.5
3
2
4
1.5
5
1
6
0.5
Diversification rates
0
18
O (ppt)
4
0
20
15
10
Million years before present
Miocene
5
0
Pliocene
Pleistocene
Fig. 8. Diversification rates calculated for the genus Impatiens (black line). Superimposed (grey line) is a time-averaged record of sea surface temperatures (Zachos et al.,
2001).
confirms the hypothesis of Grey-Wilson (1980a) suggesting an origin for Balsaminaceae ca. 50 million years ago.
It lasted another 18 million years before the lineage leading to
the extant species of Impatiens diverged from the lineage leading
to the extant species of Hydrocera (30.7 mya, Late Eocene). Despite
this early separation, the extant species of Impatiens have a coalescence point in the Early Miocene (22.5 mya; Fig. 3). It is unclear
whether this phenomenon of no net diversification during a period
of 8 million years is due to low rates of speciation or high rates of
extinction because no fossil data are available for this group. A similar question applies to the sister group of Impatiens, the monospecific Hydrocera triflora, which has a restricted Indo-Malaysian
distribution (Grey-Wilson, 1980b). Since the origin of this lineage
in the Late Eocene, climatic conditions have changed significantly,
resulting in several periods of drought and huge fluctuations in sea
level (Berggren et al., 1995; Haq et al., 1987; Riggs, 1984). Due to
its restriction to swampy lowland habitats (not occurring above
100 m altitude) and a less efficient dispersal mechanism than its
sistergroup Impatiens, other Hydrocera lineages could have gone
extinct during that time. Alternatively, speciation could be slowed
down by other causes (e.g. fluctuating sea levels), resulting in only
very low diversification. This remarkable situation in which lack of
diversification occurred in the sister lineage of a very species-rich
genus is analogous with the monospecific genus Hillebrandia and
its species-rich sister genus Begonia (ca. 1400 species, Clement
et al. 2004).
4.3. Diversification patterns in Impatiens: the effect of a global climate
change
Despite our ability to reconstruct the basal phylogenetic relationships within Impatiens using only the chloroplast atpB-rbcL
spacer data, most of the recently diversified lineages remained
unresolved. By combining the rapidly evolving ImpDEF1 and ImpDEF2 markers with the chloroplast atpB-rbcL spacer, we were able
to resolve many of the previously unresolved relationships in
Impatiens. However, due to rapid evolution of the ImpDEF genes,
we were obliged to use an approach of secondary calibration to assess the evolutionary history of the rapid diversification of Impatiens. Several studies indicate that secondary calibration can
cause considerable problems in molecular dating (Graur and Mar-
tin, 2004; Hedges and Kumar, 2003, 2004). Nevertheless, we considered this a valid approach for our study because we used the
same dating protocol, a similar phylogeny reconstruction method
and complementary taxon sampling for both dating analyses. Furthermore, in order to reduce the problem of secondary calibration,
we used additional geologic evidence together with the calibration
points based on direct and indirect fossil data.
The present study indicates that diversification in Impatiens increased since the last 4.6–5 million years (Fig. 8). This period of
exceptional diversification, during which 84% of all Impatiens lineages were established, can be linked with the global cooling that
occurred during the Pliocene and Pleistocene (Fig. 6 In addition,
when comparing the proportion of 207 Impatiens lineages (b/
d = 0) equalling the calculated c-value of 6.27 with the approximately 1000 currently described taxa, it is clear that the climatic
conditions correlate well with the rapid diversification within the
genus. This can also be seen in the fluctuating species diversification rates of the last 4.6 million years (between 0.09 and 4 species
per million year; Fig. 8), which are probably related to the glacial–
interglacial cycles that were initiated during this period (Zachos
et al., 2001). Our study demonstrates that the current species richness of Impatiens is the result of a sudden diversification boost and
did not originate via gradual accumulation of species over a long
geological period. In addition the massive species diversity in
Impatiens occurred through changing climatic conditions in the Pliocene and Pleistocene, and was not the outcome of an enhanced
speciation event in the Tertiary or earlier. Consequently, the ‘‘refuge” hypothesis, in which rainforest species populations became
isolated from each other during times of cool and dry climate conditions, may be a plausible model to explain the enormous diversity in Impatiens. The plants are characterized by explosively
dehiscent capsules with seeds travelling for the most part less than
a meter. In addition, most balsams are montane plants, often restricted to a small area, sometimes no more than a single mountain
peak or a mountain range. The periodicity of glacial cycles during
the Pleistocene could have resulted in several alternating episodes
of contraction and expansion of the montane and lowland rainforests with vegetation belts slightly shifting down and upslope during these varying climatic conditions (Kebede et al. 2007). As a
result, Impatiens would have been forced to migrate along with
the rainforest belts in order to retain its required habitat type. With
822
S.B. Janssens et al. / Molecular Phylogenetics and Evolution 52 (2009) 806–824
its area of distribution being extremely fragmented during certain
climatic episodes, many different Impatiens populations were
probably isolated for several thousands of years. The continuous
cycle of dispersal subsequent to fragmentation and isolation that
has been induced by climatic fluctuation during the Pliocene and
Pleistocene almost certainly contributed to the rapid radiation of
Impatiens.
4.4. Biogeographic patterns in Impatiens
Biogeographic reconstruction indicates that the center of origin
for Impatiens is Southwest China, from where the genus subsequently dispersed to Africa, India, the Himalayas, Southeast Asia,
Central Asia and North America. These results are consistent with
the fact that China contains 222 species or roughly 1/4th of all currently described Impatiens species, thereby largely surpassing
South India and Sri Lanka, the Himalayas, Africa (Grey-Wilson,
2008). Of these Chinese species, the majority is endemic to the
provinces of Sichuan and Yunnan in the Southwest of China. In
addition, the only fossil bearing a relation to extant Impatiens
was found in the East China Sea (I. brevicolpus – Santan Formation,
Pliocene; Song et al., 2004).
In the past, several morphological similarities have been observed in species endemic to Africa and South India, suggesting a
close affinity between these taxa and even a possible migration
route connecting these two areas (Grey-Wilson, 1980a). Despite
the resemblance between various African and South Indian balsams, the present study illustrates that the South Indian species
originated from two independent dispersal events (clades SI1 and
SI2), with Southwest China reconstructed as the area of origin for
each clade. However, one of the two South Indian lineages (clade
SI2) is closely related to one of the African subclades (clade A3),
hence explaining the overall similarities between taxa from two
clades. This African-affiliated lineage colonized South India between the Late Pliocene and the Early Pleistocene. In contrast,
the other South Indian lineage (clade SI1) previously dispersed in
the Late Miocene.
Based on the close resemblance of the North American lineage
to some Chinese species, Grey-Wilson (1980a) suggested that the
genus arrived only recently in the New World. Our data corroborate his interpretation, as the colonization of North America is estimated to have occurred in the Pleistocene between 1.32 and
1.27 million years ago (clade NA). Interestingly, this time estimate
corresponds with a glacial maximum that lasted from 1.35 until
1.25 mya (Berger et al., 1999; Bintanja and Van de Wal, 2008; Raymo 1997). The North American clade is sister to Impatiens nolitangere, one of the few Impatiens species that is not endemic to a
restricted geographic area but is a widespread species in temperate
Eurasia, including Southwest China (Chen et al., in press). North
American may have been colonized via the Bering land bridge in
association with glacial–interglacial stages during the Pleistocene.
Because some Chinese species are distributed as far as the
Himalayan chain, it has been considered that there is a strong biogeographic correlation between these two distribution areas
(Grey-Wilson, 2008). Present data show that all species endemic
to the Himalayas have a sister species which is distributed or even
endemic to China. As many other hotspots for Impatiens, also the
Himalayan region is characterized by a considerable number of
species (ca. 120 ssp.). All these Himalayan endemics appear to find
their origin in (Southwest) China via several separate dispersal
events, which occurred as early as the Late Miocene and not via
a single colonization event as for example with the Malagasy or
North American Impatiens species.
The sister species to the New Guinean endemic, Impatiens hawkeri, is located in the Southeast Asian Archipelago. In addition, the
closest relative of these two species is distributed in South China
and the northern provinces of India, Myanmar and Vietnam close
to the Chinese border. As a result, the most recent common ancestor of the New Guinean species dispersed from South China towards the Southeast Asian Archipelago from where New Guinea
was eventually reached (clade SEA1). According to our results,
the lineage towards I. hawkeri recently crossed the Wallace’s Line
from Sunda to Sahul Shelves, thereby arriving on New Guinea during the Early Pleistocene. This recent age estimate is in concordance with the difficulties that were encountered to delimit
species of the existing lineages on the island (Grey-Wilson,
1980b). As a result, Grey-Wilson lumped all previously described
Papuasian Impatiens species in a single hugely variable species,
which is subdivided in 15 groups. A population genetic approach
will be required to further investigate the complex evolution of
this group.
Ancestral area reconstruction clearly indicates that Impatiens
colonized Africa on three separate occasions. Previous studies
demonstrated that Africa was colonized more than once, but the
exact number of dispersal events could not be determined (Janssens et al., 2006; Yuan et al., 2004). The most recent common
ancestor of the three African clades is located in Southwest China.
Although our initial work (Janssens et al., 2006) left open the possibility that India may have been colonized from Africa, this expanded study suggests that African Impatiens never recolonized
Asia. The only colonization events of Impatiens from Africa were
to the offshore islands of the Seychelles and Madagascar (and then
to the Comoros) in the Indian Ocean. Although only 4% (7/ ± 150) of
the Malagasy Impatiens species were sampled for this study, all
species are closely related, raising the possibility that most or all
of this diversity was derived from a single colonization event
(clade M). The initial colonization of Africa by Impatiens apparently
occurred relatively early, as the oldest group arrived during the
Late Miocene (clade A1). All species present in this old African
clade are restricted to East and South Africa and characterized by
disjunct distribution areas. Considering the older age of this lineage, the disjunct distribution areas of its representatives and the
low rate of diversification, this clade may have experienced extinction in the intervening areas. The second dispersal of Impatiens to
Africa occurred between the Late Miocene and Early Pliocene
(clade A2). Representatives of this clade are endemic to West
Africa. The third colonization of Africa occurred during the Late
Miocene and Early Pliocene (clade A3). This lineage quickly diversified after reaching Africa, resulting in the largest diversification of
Impatiens in Africa. Most Impatiens species in this clade are restricted to the montane forests and highlands of Central and East
Africa. Additionally, only a few species are present in both Central
and West Africa. Most species that occur in both Central and West
Africa have a disjunct distribution, only occurring in mountainous
regions of West Africa, and the eastern part of Congo and neighbouring mountainous regions, yet absent in the lowland forests
of central Congo. One exception is Impatiens niamniamensis, which
has a continuous distribution from Central to West Africa throughout the Congo River basin. Because the present Congolian lowland
forests have been considerably affected by periods of glacial aridity, we assume that this species recolonized the Congo basin only
recently. Species with disjunct distribution areas in Central and
West Africa may have also been affected by ice age aridity. For
example Impatiens mannii is inferred to have originated during
the Pleistocene in the volcanic regions of Central Africa and may
have migrated to West Africa, in which case it would have been
more widespread throughout the tropical rainforests of central
Africa. Climatic fluctuations related to glacial events reduced the
central African rainforests (Coetzee, 1993; Plana, 2004). Only the
more mountainous populations of I. mannii may have survived,
having failed to recolonize the lowlands like I. niamniamensis.
Alternatively, I. mannii may have reached West Africa via long-
S.B. Janssens et al. / Molecular Phylogenetics and Evolution 52 (2009) 806–824
distance dispersal. In either case, the morphological similarity between both western and central African populations of I. mannii
indicates a recent isolation.
Acknowledgments
We thank the Tanzanian Commission for Science and Technology, the University of Dar es Salaam, the Ugandan National Research Council, Makerere University, the National Museums of
Kenya, the National Herbarium of Zimbabwe and the University
of the Witwatersrand for permission and collaboration while conducting fieldwork in the respective countries. We are also grateful
to the National Botanic Garden of Belgium (BR), National Herbarium of the Netherlands, Leiden University branch (L), Utrecht University branch (U) and Wageningen University branch (WAG),
Botanischer Garten Berlin-Dahlem (B), University of California
Botanical Garden at Berkeley (UC), Botanical Garden of Marburg
(MB), Botanical Garden of the University of Copenhagen (C), Royal
Botanic Garden Edinburgh (E), South China Botanical Garden
(IBSC), Denver Botanic Gardens (DBG), Holden Arboretum, YongMing Yuan, Yi Song, Ray Morgan, Steven Dessein and Anke Geeraerts for providing plant material. Tom Van England is acknowledged for his help with the statistical methods used in this
paper. This study was financially supported by research grants of
the K.U.Leuven (OT/05/35) and the Fund for Scientific ResearchFlanders (FWO Belgium) (G.0104.01). Steven Janssens holds a
PhD research grant from FWO.
References
Angiosperm Phylogeny Group, 2003. An update of the Angiosperm Phylogeny
Group classification for the orders and families of flowering plants: APG II. Bot.
J. Linn. Soc. 141, 399–436.
Anderson, C.L., Bremer, K., Friis, E.M., 2005. Dating phylogenetically basal eudicots
using rbcL sequences and multiple reference points. Am. J. Bot. 92, 1737–1747.
Aubréville, A., 1962. Savanisation tropicale et glaciations Quaternaires. Adansonia 2,
16–84.
Basinger, J.F., Christophel, D.C., 1985. Fossil flowers and leaves of the Ebenaceae
from the Eocene of southern Australia. Can. J. Bot. 63, 1825–1843.
Baldwin, B.G., Sanderson, M.J., 1998. Age and rate of diversification of the Hawaiian
silversword alliance (Compositae). Proc. Natl. Acad. Sci. USA 95, 9402–9406.
Bell, C.D., Donoghue, M.J., 2005. Dating the Dipsacales: comparing models, genes,
and evolutionary implications. Am. J. Bot. 92, 284–296.
Berger, A., Li, X.S., Loutre, M.F., 1999. Modeling northern hemisphere ice volume
over the last 3 Ma. Quat. Sci. Rev. 18, 1–11.
Berggren, W.A., Kent, D.V., Swisher, I.C.C., Aubry, M.-P., 1995. A revised Cenozoic
geochronology and chronostratigraphy. In: Berggren, W.A., Kent, D.V., Aubry,
M.-P., Hardenbol, J. (Eds.), Geochronology, Time Scales and Global Stratigraphic
Correlation. Society for Economic Paleontology and Mineralogy, Tulsa,
Oklahoma, pp. 129–212.
Bhaskar, V., Razi, B.A., 1981. Peninsular Indian Impatiens L. (Balsaminaceae). Bull.
Bot. Surv. India. 23, 191–196.
Bintanja, R., van de Wal, R.S.W., 2008. North American ice-sheet dynamics and the
onset of 100,000-year glacial cycles. Nature 454, 869–872.
Bremer, B., Bremer, K., Heidari, N., Erixon, P., Olmstead, R.G., Anderberg, A.A.,
Källersjö, M., Barkhordarian, E., 2002. Phylogenetics of asterids based on 3
coding and 3 non-coding DNA at higher taxonomic levels. Mol. Phylogenet. Evol.
24, 274–301.
Bremer, K., Friis, E.M., Bremer, B., 2004. Molecular phylogenetic dating of
asterid flowering plants shows Early Cretaceous diversification. Syst. Biol.
53, 496–505.
Bowie, R.C.K., Fjeldså, J., Hackett, S.J., Crowe, T.M., 2004. Molecular evolution in
space and though time: mtDNA phylogeography of the Olive Sunbird
(Nectarinia olivacea/obscura) throughout continental Africa. Mol. Phylogenet.
Evol. 33, 56–76.
Call, V.B., Dilcher, D.L., 1992. Investigations of angiosperms from the Eocene of
southeastern North America: Samaras of Fraxinus wilcoxiana. Rev. Paleobot.
Palynol. 74, 249–266.
Chen, Y.-L., 2001. Balsaminaceae. In: Chen, Y.-L. (Ed.), Flora Reipublicae Popularis
Sinicae, Tomus 47. Science Press, Beijing, pp. 1–243.
Chen,Y., Akiyama, S. Ohba, H., in press. Flora of China: Balsaminaceae.
Clement, W.L., Tebbitt, M.C., Forrest, L.L., Blair, J.E., Brouillet, L., Eriksson, T.,
Swensen, S.M., 2004. Phylogenetic position and biogeography of Hillebrandia
sandwicensis (Begoniaceae): a rare. Hawaiian relict. Am. J. Bot. 91, 905–917.
Coetzee, J.A., 1993. African flora since the terminal Jurassic. In: Goldblatt, P. (Ed.),
Biological Relationships Between Africa and South America. Yale University
Press, New Haven, pp. 37–61.
823
Collinson, M.E., Boulter, M.C., Holmes, P.L., 1993. Magnoliophyta (‘Angiospermae’).
In: Benton, M.J. (Ed.), The Fossil Record, vol. 2. Chapman and Hall, London, pp.
809–841.
Crowe, T.M., Crowe, A.A., 1982. Patterns of distribution, diversity and endemism in
Afrotropical birds. J. Zool. 198, 417–442.
Davies, T.J., Barraclough, T.G., Chase, M.W., Soltis, P.S., Soltis, D.S., Savolainen, V.,
2004. Darwin’s abominable mystery: insights from a supertree of the
angiosperms. Proc. Natl. Acad. Sci. USA 101, 1904–1909.
Dressler, S., 2004. Balsaminaceae. In: Kubitzki, K. (Ed.), Families and Genera of
Vascular Plants, vol. 6. Springer-Verlag, Berlin, Heidelberg, pp. 258–265.
Erkens, R.H.J., Chatrou, L.W., Maas, J.W., van der Niet, T., Savolainen, V.,
2007. A rapid diversification of rainforest trees (Guatteria; Annonaceae)
following dispersal from Central into South America. Mol. Phylogenet.
Evol. 44, 399–411.
Felsenstein, J., 1995. PHYLIP: Phylogeny Inference Package. Version 3.57c.
Department of Genetics, University of Washington, Seattle, Washington.
Fischer, E., 2004. Balsaminaceae. In: Kubitzki, K. (Ed.), Families and Genera of
Vascular Plants, vol. 6. Springer-Verlag, Berlin, Heidelberg, pp. 20–25.
Gamble, T., Bauer, A.M., Greenbaum, E., Jackman, T.R., 2008. Evidence for
Gondwanan vicariance in an ancient clade of gecko lizards. J. Biogeogr. 35,
88–104.
Glor, R.E., Flecker, A.S., Benard, M.F., Power, A.G., 2001. Lizard diversity and
agricultural disturbance in a Caribbean forest landscape. Biodiv. Conserv. 10,
711–723.
Graham, A., 1975. Late Cenozoic evolution of tropical lowland vegetation in
Veracruz, Mexico. Evolution 29, 723–735.
Graur, D., Martin, W., 2004. Reading the entrails of chickens: molecular timescales
of evolution and the illusion of precision. Trends Genet. 20, 80–86.
Grey-Wilson, C., 1980a. Impatiens of Africa. Balkema, Rotterdam.
Grey-Wilson, C., 1980b. Hydrocera triflora, its floral morphology and relationship
with Impatiens. Studies in Balsaminaceae: V. Kew Bull. 35, 213–219.
Grey-Wilson, C., 1989. A revision of Sumatran Impatiens. Studies in Balsaminaceae:
VIII. Kew Bull. 44, 107–113.
Grey-Wilson, C., 2008. Impatiens forrestii: Balsaminaceae. Curtis’s Bot. Mag. 25, 139–
145.
Haffer, J., 1969. Speciation in Amazonian forest birds. Science 165, 131–137.
Haffer, J., 1982. General aspects of the refuge theory. In: Prance, G.T. (Ed.), Biological
Diversification in the Tropics. Columbia University Press, pp. 6–24.
Haffer, J., Prance, G.T., 2001. Climatic forcing of evolution in Amazonia during the
Cenozoic: on the refuge theory of biotic differentiation. Amazoniana 16, 579–
607.
Haq, B.U., Hardenbol, J., Vail, P.R., 1987. Chronology of fluctuating sea levels since
the Triassic. Science 235, 1156–1167.
Harris, D.J., Poulsen, A.D., Frimodt-Möller, C., Preston, J., Cronk, Q.C.B., 2000. Rapid
radiation in Aframomum (Zingiberaceae): evidence from nuclear ribosomal DNA
internal transcribed spacer (ITS) sequences. Edinburgh J. Bot. 57, 377–395.
Hedges, S.B., Kumar, S., 2003. Genomic clocks and evolutionary timescales. Trends
Genet. 19, 200–206.
Hedges, S.B., Kumar, S., 2004. Precision of molecular time estimates. Trends Genet.
20, 242–247.
Hughes, C., Eastwood, R., 2006. Island radiation on a continental scale: exceptional
rates of plant diversification after uplift of the Andes. Proc. Natl. Acad. Sci. USA
103, 10334–10339.
Janssens, S.B., Knox, E.B., Dessein, S., Smets. E.F., 2009. Impatiens msisimwanensis
(Balsaminaceae): Description, pollen morphology and phylogenetic positioning
of a new East African species. S. Afr. J. Bot. 75, 104-109.
Janssens, S., Geuten, K., Yuan, Y.-M., Song, Y., Küpfer, P., Smets, E., 2006.
Phylogenetics of Impatiens and Hydrocera (Balsaminaceae) using chloroplast
atpB-rbcL spacer sequences. Syst. Bot. 31, 171–180.
Janssens, S., Geuten, K., Viaene, T., Yuan, Y.-M., Song, Y., Smets, E., 2007.
Phylogenetic utility of the AP3/DEF K-domain and its molecular evolution in
Impatiens (Balsaminaceae). Mol. Phylogenet. Evol. 43, 225–239.
Jaramillo, C., Rueda, M.J., Mora, G., 2006. Cenozoic plant diversity in the Neotropics.
Science 311, 1893–1896.
Johnson, W.E., Eizirik, E., Pecon-Slattery, J., Murphy, W.J., Antunes, A., Teeling, E.,
O’Brien, S.J., 2006. The late Miocene radiation of modern Felidae: a genetic
assessment. Science 311, 73–77.
Kebede, M., Ehrich, D., Taberlet, P., Nemomissa, S., Brochmann, C., 2007.
Phylogeography and conservation genetics of a giant lobelia (Lobelia giberroa)
in Ethiopian and Tropical East African mountains. Mol. Ecol. 16, 1233–1243.
Kim, K.-J., Choi, K.S., Jansen, R.K., 2005. Two chloroplast DNA inversions originated
simultaneously during the early evolution of the sunflower family (Asteraceae).
Mol. Biol. Evol. 22, 1783–1792.
Knapp, S., Mallet, J., 2003. Refuting Refugia? Science 300, 71–72.
Knox, E.B., Palmer, J.D., 1998. Chloroplast DNA evidence on the origin and radiation
of the giant lobelias in eastern Africa. Syst. Bot. 23, 109–149.
Lancucka-Srodoniowa, M., 1967. Two new genera: Hemiptelea Planch. and Weigela
Thurnb. in the younger Tertiary of Poland. Acta Palaeobot. 8, 1–17.
Leopold, E.B., 1969. Miocene pollen and spore flora of Eniwetok, Marshall Islands.
US Geol. Surv. Profess. Pap. 260, 1183–1185.
Maddison, W.P., Maddison, D.R., 2008. Mesquite: a modular system for evolutionary
analysis. Version 2.5. Available from: <http://mesquiteproject.org>.
Maddison, D., Maddison, W., 2002. MacClade. Sinauer Associates, Sunderland,
Massachusetts.
Maley, J., 1987. Fragmentation de la fôret dense humide africaine et extension des
biotopes montagnards au Quaternaire recent: nouvelles données polliniques et
824
S.B. Janssens et al. / Molecular Phylogenetics and Evolution 52 (2009) 806–824
chronologiques. Implications paleoclimatiques et biogeographiques. In:
Coetzee, J.A. (Ed.), Palaeoecology of Africa and the Surrounding Islands, vol.
18. Balkema, Rotterdam, pp. 307–334.
Mayr, E., O’Hara, R.J., 1986. The biogeographic evidence supporting the Pleistocene
forest refuge hypothesis. Evolution 40, 55–67.
Merckx, V., Chatrou, L.W., Lemaire, B., Sainge, M.N., Huysmans, S., Smets, E.F., 2008.
Diversification of myco-heterotrophic angiosperms: Evidence from
Burmanniaceae. BMC Evol. Biol. 8, 178–194.
Moritz, C., Patton, J.L., Schneider, C.J., Smith, T.B., 2000. Diversification of
rainforest faunas: an integrated molecular approach. Ann. Rev. Ecol. Syst.
31, 533–563.
Muller, J., 1981. Fossil pollen records of extant angiosperms. Bot. Rev. 47, 1–142.
Near, T.J., Sanderson, M.B., 2004. Assessing the quality of molecular divergence time
estimates by fossil calibrations and fossil-based model selection. Philos. Trans.
R. Soc. Ser. B 359, 1477–1483.
Pagel, M., 1999. The maximum likelihood approach to reconstructing ancestral
character states of discrete characters on phylogenies. Syst. Biol. 48, 612–622.
Pennington, R.T., Lavin, M., Prado, D.E., Pendry, C.A., Pell, S.K., Butterworth, C.A.,
2004. Historical climate change and speciation: neotropical seasonally dry
forest plants show patterns of both Tertiary and Quaternary diversification.
Philos. Trans. R. Soc. Ser. B 359, 515–537.
Plana, V., Gascoigne, A., Forrest, L.L., Harris, D., Pennington, R.T., 2004. Pleistocene
and pre-Pleistocene Begonia speciation in Africa. Mol. Phylogenet. Evol. 31, 449–
461.
Plana, V., 2004. Mechanisms and tempo of evolution in the African Guineo–
Congolian rainforest. Philos. Trans. R. Soc. Lond. B. 359, 1585–1594.
Pòcs, T., 2007. A new species of Impatiens (Balsaminaceae) from the Nguru
Mountains (Tanzania). Acta Bot. Hung. 49, 377–383.
Prance, G.T., 1973. Phytogeographic support for the theory of Pleistocene forest
refuges in the Amazon Basin, based upon evidence from distribution pattern in
Caryocaraceae, Chrysobalanaceae, Dichapetalaceae and Lecythidaceae. Acta
Amazon. 3, 5–28.
Prance, G.T., 1982. A review of the phytogeographic evidences for Pleistocene
climate changes in the Neotropics. Ann. Mo. Bot. Gard. 69, 594–624.
Posada, D., Crandall, K.A., 1998. Modeltest, testing the model of DNA substitution.
Bioinformatics 14, 817–818.
Pybus, O.G., Rambaut, A., 2002. GENIE: estimating demographic history from
molecular phylogenies. Bioinformatics 18, 1404–1405.
Pybus, O.G., Harvey, P.H., 2000. Testing macro-evolutionary models using
incomplete molecular phylogenies. Philos. Trans. R. Soc. Lond. B 267, 2267–
2272.
Rahelivololona, M.E., Fischer, E., Wohlhauser, S., 2003. Balsaminaceae. In: Goodman,
S.M., Benstead, J.P. (Eds.), The Natural History of Madagascar. The University of
Chicago Press, Chicago and London, pp. 402–409.
Rambaut, A., 2002. PHYLOGEN: Phylogenetic Tree Simulator Package. University of
Oxford, Oxford.
Rambaut, A., Charleston., M., 2002. TreeEdit: Phylogenetic Tree Editor v. 1.0 alpha
10. Available from: <http://evolve.zoo.ox.ac.uk/software/TreeEdit>.
Raymo, M.E., 1997. The timing of major climate terminations. Paleoceanography 12,
577–585.
Reid, E.M., Chandler, M.E.J., 1926. In: Catalogue of Cainozoic plants in the
Department of Geology, vol. 1. The Bembridge Flora. Br. Mus. Nat. Hist.,
London, p. 206.
Richardson, J.E., Pennington, R.T., Pennington, T.D., Hollingsworth, P.M., 2001. Rapid
diversification of a species-rich genus of neotropical rain forest trees. Science
293, 2242–2245.
Riggs, S.R., 1984. Paleoceanographic model of Neogene phosphorite deposition, US
Atlantic Continental Margin. Science 223, 123–131.
Rull, V., 1999. Palaeofloristic and palaeovegetational changes across the Paleocene–
Eocene boundary in northern South America. Rev. Palaeobot. Palynol. 107, 83–
95.
Sanderson, M.J., 2002. Estimating absolute rates of molecular evolution and
divergence times: a penalized likelihood approach. Mol. Biol. Evol. 19, 101–109.
Savolainen, V., Chase, M.W., Hoot, S.B., Morton, C.M., Soltis, D.E., Bayer, C., Fay, M.F.,
De Bruijn, A., Sullivan, S., Qiu, Y.-L., 2000. Phylogenetics of flowering plants
based upon a combined analysis of plastid atpB and rbcL gene sequences. Syst.
Biol. 49, 306–362.
Schluter, D., Price, T., Mooers, A.O., Ludwig, D., 1997. Likelihood of ancestral states in
adaptive radiation. Evolution 51, 1699–1711.
Schönenberger, J., Anderberg, A.A., Sytsma, K.J., 2005. Molecular phylogenetics and
patterns of floral evolution in the Ericales. Int. J. Plant Sci. 166, 265–288.
Shimizu, T., 1979. The lectures on the flora of Thailand. Acta Phytotax. Geobot. 30,
180–187.
Song, Z.-C., Wang, W.-M., Huang, F., 2004. Fossil pollen records of extant
angiosperms in China. Bot. Rev. 70, 425–458.
Stamatakis, A., 2006. RAxML-VI-HPC: maximum likelihood-based phylogenetic
analyses with thousands of taxa and mixed models. Bioinformatics 22, 2688–
2690.
Stamatakis, A., Ludwig, T., Meier, H., 2005. Raxml-III: a fast program for maximum
likelihood-based inference of large phylogenetic trees. Bioinformatics 21, 456–
463.
Swofford, D.L., 2002. PAUP. Phylogenetic Analysis Using Parsimony (and
other
Methods).
Version
4.
Sinauer
Associates,
Sunderland,
Massachusetts.
Takahashi, M., Crane, P.R., Manchester, S.R., 2002. Hironoia fusiformis gen. et sp.
nov.; a cornalean fruit from the Kamikitaba locality (Upper Cretaceous, Lower
Coniacian) in northeastern Japan. J. Plant Res. 115, 463–473.
Toppin, S.M., 1920. Notes on the balsams of Chitral and the Kachin Hills. Bull. Misc.
Inf. Kew 34, 5–366.
Wiens, J.J., 2005. Missing data and the design of phylogenetic analyses. J. Biomed.
Inform. 39, 34–42.
Yuan, Y.-M., Song, Y., Geuten, K., Rahelivololona, E., Wohlhauser, S., Fischer, E.,
Smets, E., Küpfer, P., 2004. Phylogeny and biogeography of Balsaminaceae
inferred from ITS sequence data. Taxon 53, 391–403.
Zachos, J., Pagani, M., Sloan, L., Thomas, E., Billups, K., 2001. Trends,
rhythms, and aberrations in global climate 65 Ma to present. Science
292, 686–693.
Zink, R.M., Slowinsky, J.B., 1995. Evidence from molecular systematics for decreased
avian diversification in the Pleistocene epoch. Proc. Natl. Acad. Sci. USA 92,
5832–5835.