Klitgrd2013 PDF

Download as pdf or txt
Download as pdf or txt
You are on page 1of 15
 
A detailed investigation of the Pterocarpus clade (Leguminosae: Dalbergieae):
Etaballia
 with radially symmetrical
 󿬂
owers is nested within thepapilionoid-
󿬂
owered
 Pterocarpus
Bente B. Klitgård
, Félix Forest
b
, Thomas J. Booth
b
, C. Haris Saslis-Lagoudakis
b,c
a
Herbarium, Library, Art and Archives, Royal Botanic Gardens, Kew, Richmond, TW93AE, United Kingdom
b
 Jodrell Laboratory, Royal Botanic Gardens, Kew, Richmond, TW93DS, United Kingdom
c
Macroevolution and Macroecology Group, Research School of Biology, Australian National University, Canberra, 0200, Australia
a b s t r a c ta r t i c l e i n f o
Available online 28 August 2013Edited by JC Manning
Keywords:
Phylogeny
EtaballiaPterocarpus
Radially symmetric
 󿬂
owersITS2trnL-FndhF-rpL32matKrbcL 
The pantropical genus
 Pterocarpus
 (Leguminosae: Dalbergieae) with papilionoid
 󿬂
owers, and allied gen-era in the Pterocarpus clade were sampled for the
 󿬁
ve molecular markers ITS2,
 trnL-F 
,
 ndhF-rpL32
,
 matK 
,and
 rbcL
, as part of our ongoing systematic studies in the clade. For wider analyses of the Pterocarpusclade the remaining 14 members of this clade were also sampled for
 matK 
. Phylogenetic analyses wereperformed under the maximum likelihood criterion (ML) and Bayesian criteria. In the
 󿬁
ve-marker anal-ysis of the core Pterocarpus clade (including 106 accessions) two robustly supported clades were re-solved. The
 󿬁
rst clade includes
 Centrolobium
,
 Etaballia
,
 Inocarpus
,
 Maraniona
,
 Paramachaerium
,
Pterocarpus
,
 Ramorinoa
, and
 Tipuana
. The second includes all species of 
 Pterocarpus
 (except
 P.acapulcensis
),
 Etaballia
 with radially symmetric
 󿬂
owers, and
 Paramachaerium
.
 Paramachaerium
 is placedas sister to the several
 Pterocarpus
 species from South America, while
 Etaballia
 is resolved within theclade containing the African and Asian species of 
 Pterocarpus
. The wider sampled
 matK 
 data set includes199 accessions.
 Discolobium
 and
 Riedeliella
 are recovered as sister to the remaining Pterocarpus clade.
Platymiscium
 is strongly supported as sister to the rest of the members of the clade, and
 Pterocarpusacapulcensis
 is also here resolved in a separate lineage from the remaining
 Pterocarpus
 accessions. Weused the phylogenies to investigate patterns of 
 󿬂
oral evolvability in the Pterocarpus clade, which includefour genera with actinomorphic
 󿬂
owers (
 Acosmium s.s.
,
 Etaballia
,
 Inocarpus
 and
 Riedelliela
). Our results re-inforce the hypothesis that
 󿬂
ower evolvability is high in early-branching legume lineages, and that actino-morphy has evolved independently four times in the Pterocarpus clade. In light of our results, thetaxonomic status of the monospeci
󿬁
c genus
 Etaballia dubia
 Benth. was revisited, and the species is synon-ymized as belonging to
 Pterocarpus
, under the name
 Pterocarpus dubius
 Spreng., published in 1827, buthiding in synonymy for nearly two centuries.Crown Copyright © 2013 Published by Elsevier B.V. on behalf of SAAB. All rights reserved.
1. Introduction
The decrease in costs and increase in methodological ef 
󿬁
ciencyin molecular techniques in the past decade have resulted in an ex-plosion in molecular phylogenetic research in general, includingstudies focused on the legume family. Among
 󿬂
owering plant fam-ilies, Leguminosae encompasses one of the widest ranges in geo-graphic distribution, habitat preference, habit type, leaf,
 󿬂
owerand fruit morphology, thus the family serves as an excellentmodel for evolutionary studies [e.g. Dalbergioids (Lavin et al.,2001),
 Aeschynomene
 Acosmium
 Lonchocarpus
 Steinbachiella
(Lewis et al., 2012)]. Phylogenetic studies within the Leguminosaehave not only contributed to a much better understanding of rela-tionships among legume genera, but they have also sometimeshighlighted cases where a recircumscription of the generic and/ortribal concepts was necessary [e.g. Dalbergioids (Lavin et al.,2001),
 Aeschynomene
 Acosmium
 Lonchocarpus
 Steinbachiella
(Lewis et al., 2012)]. Similarly, these studies have often required arethink of widely accepted views about habitat preference and mor-phological evolvability (e.g. Cardoso et al., 2012a,b,c; Lavin et al.,2001; McMahon and Hufford, 2004; Pennington et al., 2000;
 Corresponding author. Tel.: + 44(0)2083325257.
E-mail address:
 B.Klitgaard@kew.org (B.B. Klitgård).0254-6299/$
 –
 see front matter. Crown Copyright © 2013 Published by Elsevier B.V. on behalf of SAAB. All rights reserved.http://dx.doi.org/10.1016/j.sajb.2013.07.006
Contents lists available at ScienceDirect
South African Journal of Botany
 journal homepage: www.elsevier.com/locate/sajb
 
󿬁
rst drew attention to 23 papilionoid genera with radially sym-metric
 󿬂
owers, representing nine reversals from zygomorphy to ac-tinomorphy in their phylogenetic analyses of early-divergingpapilionoids. This number increased to twelve with the additionalthree discovered subsequently by Lavin et al. (2001) within the rel-atively early-branching Dalbergioid clade. These taxa were thoughtto be
 
primitive
 by authors of earlier legume classi
󿬁
cations (e.g.Polhill, 1981, 1994; Yakovlev, 1975), but showed to be scatteredamong early-branching papilionoid lineages including mainly gen-era with
 
less primitive
 papilionoid
 󿬂
owers (Lavin et al., 2001;Pennington et al., 2000, 2001). Those authors also noted that de-spite apparent multiple origins of actinomorphic
 󿬂
owers in theselineages, the non-papilionoid
 󿬂
ower morphologies do not occur inmore derived lineages of the Papilionoideae. Explaining this phylo-genetic pattern in the distribution of actinomorphic
 󿬂
owers,Pennington and co-authors (2000) suggested that in groups wherepapilionoid
 󿬂
owers have become more canalized (e.g. a zygomor-phic corolla with fusion of petals and stamens, often seen in themore derived legume taxa such as
 Medicago
 spp.,
 Pisum
 spp., and
Swainsona
 spp.), it is more dif 
󿬁
cult to revert to a non-papilionoid
󿬂
ower morphology. Furthermore, zygomorphy might be conservedin late-branching lineages, as it has been a key innovation associat-ed with high diversi
󿬁
cation rates in various lineages (e.g. Arroyo,1981; Citerne et al., 2010; Endress, 1999, 2001; Sargent, 2004). Re-cently, this evolutionary pattern of repeated reversal from zygo- toactinomorphy has proven even more widespread among theearly-branching papilionoids (Boatwright et al., 2008; Cardoso etal., 2012a, 2012b, 2013).Since Bronn ex De Candolle (1825) coined the tribe nameDalbergieae it has taken several shapes and guises until Lavin etal.'s (2001) groundbreaking paper united 44 genera, previouslythought to belong to
 󿬁
ve different tribes, in one monophyleticgroup. They called this expanded group the
 
Dalbergioid clade
,and subdivided it into three well-supported subclades: theAdesmia, Pterocarpus, and Dalbergia clades. This paper alsoformed the basis for the circumscription of the Dalbergieae(Klitgård and Lavin, 2005) in Legumes of the World (Lewis et al., 2005). Since 2005, three new genera have been added to thegroup, increasing the number of genera assigned to the Dalbergieaeto a current total of 47 members.
 Steinbachiella
 Harms was reinstat-ed and placed in the Dalbergia clade (Lewis et al., 2012), while thePterocarpus clade acquired two new additions: the monospeci
󿬁
c
Maraniona lavinii
 C.E. Hughes, G.P. Lewis, Daza & Reynel., a new dis-covery from the interandean valleys of Peru (Hughes et al., 2004),and
 Acosmium
 Schott
 s.s.
 (three spp.) joining the Pterocarpusclade segregated from the other members of 
 Acosmium
 which areplaced in the Bowdichia clade of the Genistioids under the new ge-neric names
 Leptolobium
 and
 Guianodendron
 (Cardoso et al.,2012a). In the Pterocarpus clade a number of genera have beenthe focus of recent species level phylogenetic studies:
 Platymiscium
 Centrolobium
 Pterocarpus
 Acosmium
 (Cardoso et al., 2012a), which have improved the resolu-tion of the core Pterocarpus clade. Nevertheless, infrageneric rela-tionships within the wider Pterocarpus clade are still notwell-resolved.
Etaballia
 Benth. is a monospeci
󿬁
c legume tree genus, restrictedto periodically inundated humid tropical rainforests of Brazil, theGuianas, and Venezuela, and has recently also been recorded fromPando in the Bolivian Amazon (voucher
 Klitgård et al. 1394
2). Largely due to its unusual morphology with unifoliolateleaves, radially symmetric
 󿬂
owers and samaroid fruits,
 Etaballia
has had a turbulent taxonomic history, since Bentham (1840a)
󿬁
rst published
 Etaballia guianensis
 Benth. [=
E. dubia
 (Kunth)Rudd] as a member of the Bauhinieae,
 
allied to [the genus]
Schnella
…”
. In Genera Plantarum (Bentham, 1865),
 Etaballia
 wasplaced in synonymy under
 Inocarpus
 Forst., a Southeast Asianmonospeci
󿬁
c genus also with unifoliolate leaves and radially sym-metric
 󿬂
owers, but with drupaceous fruits, and placed as
 
genusanomalum
 in tribe Dalbergieae. Subsequently, most authors havegiven
 Etaballia
 independent generic status in the tribe Dalbergieae(Dalla Torre and Harms, 1900
1907; Ducke, 1949; Funch andSantos, 1997; Klitgård and Lavin, 2005; Lavin et al., 2001; Polhill,1981), while some have placed it elsewhere in the legume family:intermediate between subfamilies Mimosoideae and Papilionoideae(Kuhlmann, 1949); in a
 
primitive position
 in tribe Sophoreae orCadieae (Hutchinson, 1964
 
 in synonymy under
 Inocarpus
; Polhill,1994; Yakovlev, 1975); and between subfamilies Caesalpinioideae andPapilionoideae (Barroso et al., 1984). Some authors have commentedon the resemblance of the fruits of 
 Etaballia dubia
 to those of thetwo amphiatlantic species,
 Pterocarpus of 
 󿬁
cinalis
 Jacq. [e.g. Kunth,1824
 —
 when publishing
 Hecastaphyllum dubia
 Kunth (=
E. dubia
);Rudd, 1970], and
 P. santalinoides
 DC. (Rojo, 1972), which share asamaroid (sometimes appearing drupaceous) fruit with a narrowwing encircling 1/3 to 2/3 of the circumference of the seed chamber(Fig. 2).Another member of the Pterocarpus clade which has had acomplex taxonomic history is
 Paramachaerium
 Ducke, currentlyconsisting of 
 󿬁
ve species distributed in the Amazonian regions of the Guianas, Venezuela, Brazil, Peru, and Panama (Klitgård andLavin, 2005; Rudd, 1981). In 1925 Ducke segregated
 Machaeriumschomburgkii
 Benth. (Bentham, 1840b) [synonymous with
Pterocarpus kuhlmannii
 Ducke (Ducke, 1922)] from
 Machaerium
,published the genus
 Paramachaerium
 and based it on the typespecies
 Paramachaerium schomburgkii
 (Benth.) Ducke. In 1935he added
 Ducke [=
Pterocarpusormosioides
 Ducke (Ducke, 1922)]. To-date three additional
Fig. 1.
 Pterocarpus dubius
 Spreng. (=
 Etaballia dubia
 (Kunth) Rudd [as
 Etaballia guianensis
 Benth.]), plates
 󿬁
rst published in Hooker's Icones (1842).129
B.B. Klitgård et al. / South African Journal of Botany 89 (2013) 128
142
5
 
species have been added:
 Paramachaerium gruberi
 Brizicky(Brizicky, 1960),
 P. krukovii
 Rudd and
 P. schunkei
 Rudd (Rudd,1981). It is therefore evident that there has been a close connec-tion between
 Paramachaerium
 and
 Pterocarpus
 already from itsinception. Although Lavin et al. (2001) resolved
 Paramachaerium
as closely related to
 Pterocarpus
, its exact systematic position hasremained unclear.Here we present an updated phylogeny of the Pterocarpusclade with higher resolution for some critical nodes, and a taxo-nomic evaluation of the monospeci
󿬁
c genus
 Etaballia
. The pres-ent study investigates intrageneric relationships and patterns of 
󿬂
oral evolvability in the Pterocarpus clade of the tribeDalbergieae, which currently includes four genera with actino-morphic
 󿬂
owers (
 Acosmium s.s.
,
 Etaballia
,
 Inocarpus
 and
 Riedelliela
).Our results further reinforce the hypothesis that
 󿬂
owerevolvability is high in early-branching legume lineages, and thatactinomorphyhasevolvedindependentlyfourtimesinthePterocarpusclade.
2. Material and methods
 2.1. Taxon sampling 
We included all
 Pterocarpus
 species as in our previous phyloge-netic study (Saslis-Lagoudakis et al., 2011), as well as eight moreaccessions of American
 Pterocarpus
 species belonging to the
P. rohrii
 complex, currently under study by Mans
󿬁
eld-Williams &Hawkins at Reading University, and Klitgård & Saslis-Lagoudakisat RBG, Kew. For
 Etaballia
 and
 Paramachaerium
, we included theaccessions available from a previous study (Lavin et al., 2001)and extended the sampling by including three additional samplesfrom both genera. Outgroup selection was based on previous
Fig.2.
Pterocarpusdubius
Spreng.A.Floweringbranchof[
B.B.Klitgård1394
etal.(USZ)],photograph©B.Klitgård,RBG,Kew.B.Fruitsof 
Pterocarpusdubius
[
L.William1308
(F)].C.Fruitsof 
Pterocarpus of 
 󿬁
cinalis
 [
B.V. Rabelo et al. 2071
 (K)]. D. Fruits of 
 Pterocarpus santalinoides
 [
D. Zappi et al. 1100
 (K)].
Fig. 3.
 Synopsis of the phylogenetic tree recovered from the maximum likelihood (ML)analysis, using all DNA markers (nrITS2,
 rbcL
,
 matK 
,
 trnL
 and
 ndhF-rpL32
). Numbersabove branches show bootstrap percentages (BP). Branches with BP
 b
 50 werecollapsed.130
 B.B. Klitgård et al. / South African Journal of Botany 89 (2013) 128
142
5
 
phylogenetic studies of the
 Pterocarpus
 clade (Hughes et al., 2004;Lavin et al., 2001; Pirie et al., 2009; Saslis-Lagoudakis et al., 2008,2011), ensuring that exemplar sequences from all major lineageswithin the clade were represented.
 Acosmium
, recently identi
󿬁
edas a member of the clade (Cardoso et al., 2012a), was also includ-ed. Our sampling includes
 󿬁
ve additional accessions for
 Maraniona
and one for
 Tipuana
.
 Discolobium
 was used as outgroup taxon,representing the
 Discolobium
Riedeliella
 clade. Voucher informa-tion is shown in Appendix 1. Additionally, we collated all
 matK 
 se-quences that are publicly available in GenBank for the
 Pterocarpus
clade from previous studies, and added our newly generated
 matK 
sequences to that dataset in order to produce a phylogenetic treewith wider and denser sampling across the clade. Exemplar speciesfrom the
 Dalbergia
 and
 Adesmia
 clades were also included, and
 Amicia
 and
 Adesmia
 were used as outgroup taxa. Taxon samplingfor this analysis is shown in Appendix 2.
 2.2. Molecular methods and phylogenetic analyses
Total DNA was extracted from 0.2 to 0.3 g of leaf and/or
 󿬂
owertissue from herbarium or silica gel dried material using a modi
󿬁
ca-tion of the Doyle and Doyle method (Csiba and Powell, 2006; Doyleand Doyle, 1987). DNA was puri
󿬁
ed using QIAquick PCR columns(Qiagen, Crawley, West Sussex, UK) following the manufacturer'sprotocol.Five DNA markers were ampli
󿬁
ed, one nuclear (internal tran-scribed spacer 2, ITS2) and four plastid (
trnL
-
 intergenic spacer,the barcoding fragment of 
 matK 
, the
 󿬁
rst half of 
 rbcL
, and the
Clade A
P. santalinus P. michelianus P. ternatus P. erinaceus P. zehntnerii P. amphymenium P. violaceus Maraniona lavinii P. lucens antunesii P. indicus Etaballia dubia P. angolensis P. indicus P. rohrii P. mildbraedii mildbraedii P. santalinoides P. santalinoides P. reticulatus P. lucens lucens P. rohrii P. hayesii P. tessmannii P. rotundifolius polyanthus P. violaceus angustifolius Paramachaerium schomburgkii P. marsupium P. soyauxii P. tinctorius chrysothrix Maraniona lavinii P. floribundus P. rohrii P. brenanii Etaballia dubia P. dalbergioides P. cf. erinaceus P. rotundifolius rotundifolius P. lucens lucens Maraniona lavinii Pterocarpus acapulcensis P. officinalis officinalis P. amphymenium P. tinctorius P. tinctorius odoratus P. osun Tipuana tipu Ramorinoa P. rohrii P. tinctorius chrysothrix Centrolobium P. lucens lucens P. monophyllus P. santalinoides P. soyauxii P. officinalis gilletii Paramachaerium schomburgkii P. tinctorius albopubescens P. floribundus P. angolensis P. rotundifolius polyanthus martinii P. indicus P. amphymenium P. amazonum P. rohrii P. santalinus P. santalinoides P. brenanii P. macrocarpus P. rotundifolius polyanthus martinii P. santalinoides P. tinctorius P. marsupium Maraniona lavinii Maraniona lavinii P. amazonum Maraniona lavinii P. erinaceus Etaballia dubia P. magnicarpus P. amphymenium Tipuana tipu Inocarpus fagifer P. steinbachianus P. villosus Paramachaerium ormosoides Etaballia dubia P. mildbreadii usambarensis P. tinctorius holtzii P. hayesii P. reticulatus P. marsupium P. steinbachianus P. santalinus Pterocarpus acapulcensis P. officinalis officinalis Paramachaerium schomburgkii P. reticulatus 
10067925780666990995410090551009892947998671009498100826210079945310072831009185985610010010056529294100100609390701008665100631009110010090789910010098605461
Clade B
531.000.971.001.001.001.001.001.001.001.001.001.001.001.000.981.000.960.951.001.001.000.991.001.001.001.001.001.001.001.001.000.990.941.001.001.000.980.931.000.990.911.001.001.001.001.001.001.001.001.001.001.00
Neotropical
Pterocarpus 
African
Pterocarpus 
Asian
Pterocarpus Paramachaerium Etaballia Pterocarpus acapulcensis 
Fig. 4.
 Phylogenetic tree recovered from the maximum likelihood (ML) analysis, using all DNA markers (nrITS2,
 rbcL
,
 matK 
,
 trnL
 and
 ndhF-rpL32
). Numbers above branches showbootstrap percentages (BP) and those below branches show posterior probabilities (PP) from a Bayesian analysis of the same dataset. Only values above 50 (BP) and 0.90 (PP) areshown. Branches with both BP
 b
 50 and PP
 b
 0.90 were collapsed.131
B.B. Klitgård et al. / South African Journal of Botany 89 (2013) 128
142
5
5
5
5
5
5
5
5
5
5
5
5

Reward Your Curiosity

Everything you want to read.
Anytime. Anywhere. Any device.
No Commitment. Cancel anytime.
576648e32a3d8b82ca71961b7a986505