ArticlePDF Available

Phylogenetic Analysis of the African Genus Gilbertiodendron J. Léonard and Related Genera (Leguminosae-Caesalpinioideae-Detarieae)

Authors:

Abstract and Figures

Premise of research: Gilbertiodendron is a genus endemic to Africa with ∼30 species made up of trees of primary dry-land, riverine, and gallery forests. Recently, the west and central African monotypic genus Pellegriniodendron was merged into Gilbertiodendron. Gilbertiodendron is one of 17 genera that form the exclusively African Berlinia clade, and this study presents the findings of a phylogenetic analysis designed to evaluate the generic limits of Gilbertiodendron and its relationships within the Berlinia clade.
Content may be subject to copyright.
1
Thursday Aug 28 2014 10:26 AM/IJPS42308/2014/175/8/jharrell/islesskitain/ritterd//ms review complete/1002/use-graphics/narrow/
default/
Int. J. Plant Sci. 175(8):000–000. 2014.
2014 by The University of Chicago. All rights reserved.
1058-5893/2014/17508-00XX$15.00 DOI: 10.1086/677648
PHYLOGENETIC ANALYSIS OF THE AFRICAN GENUS GILBERTIODENDRON J. LE
´ONARD
AND RELATED GENERA (LEGUMINOSAE-CAESALPINIOIDEAE-DETARIEAE)
Manuel de la Estrella,
1,
*
,
† Jan J. Wieringa,‡ Barbara Mackinder,§ Xander van der Burgt,§ Juan A. Devesa,† and Anne Bruneau*
*Institut de Recherche En Biologie Ve´ge´tale and De´ partement de Sciences Biologiques, Universite´ de Montre´ al, 4101 Sherbrooke Est, Que´bec
H1X 2B2, Montre´al, Canada; †Departamento de Bota´nica, Ecologı´a y Fisiologı´a Vegetal, Facultad de Ciencias, Campus de Rabanales,
Universidad de Co´rdoba, 14071 Co´ rdoba, Spain; ‡Naturalis Biodiversity Centre, Botany section, Darwinweg 2, 2333 CC Leiden,
The Netherlands; and §Herbarium, Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3AE, United Kingdom
Premise of research.Gilbertiodendron is a genus endemic to Africa with 30 species made up of trees
of primary dry-land, riverine, and gallery forests. Recently, the west and central African monotypic genus
Pellegriniodendron was merged into Gilbertiodendron.Gilbertiodendron is one of 17 genera that form the
exclusively African Berlinia clade, and this study presents the findings of a phylogenetic analysis designed to
evaluate the generic limits of Gilbertiodendron and its relationships within the Berlinia clade.
Methodology. To test the monophyly of Gilbertiodendron and its relationships with other genera, we
analyzed nucleotide sequence data from the nuclear ribosomal internal transcribed spacer and the plastid trnL
intron and trnL-F intergenic spacer, using parsimony and Bayesian analyses.
Pivotal results.Gilbertiodendron is recovered as monophyletic, including all the samples previously rec-
ognized as Pellegriniodendron diphyllum.
Conclusions. The placement of Pellegriniodendron in synonymy with Gilbertiodendron is supported by
our results. Our analyses suggest that G. diphyllum is the same taxon on both sides of the Dahomey Gap.
The G. ogoouense complex is a monophyletic group of species that needs a new taxonomic framework and
within which several new species will be described. The phylogenetic framework presentedhere and the ongoing
taxonomic revision should provide the baseline data required for adequate assessment of this group of tree
species, of which only eight have been assessed under the International Union for Conservation of Nature’s
Red List criteria.
Keywords: Didelotia, Fabaceae, ITS, Librevillea,Plagiosiphon,Pellegriniodendron, phylogenetic analyses,
tropical Africa, trnL-trnF.
Introduction
Leguminosae is the third-largest flowering plant family,
made up 19,500 species in 751 genera, occurring in a great
variety of habitats from rain forests and mangrove swamps to
deserts and temperate zones (Lewis et al. 2005; LPWG 2013a).
The family is traditionally divided into 3 subfamilies—Papil-
ionoideae, Mimosoideae, and Caesalpinioideae—but propos-
als for a new classification currently being discussed will in-
crease that number to 6, 10–12, or even 15 (LPWG 2013b).
The caesalpinioid legumes form the smallest of the three tra-
ditionally recognized subfamilies and includes 2,250 species
assigned to 171 genera and four tribes (Lewis et al. 2005;
LPWG 2013a). In terms of species richness, Leguminosae is
the most important angiosperm family in tropical Africa (Le-
brun and Stork 1998). The dominance of Caesalpinioideae
species in tropical Africa was recognized by Letouzey (1968),
who named a specific forest type, the foreˆt biafre´enne.Cae-
salpinioideae species can form large expanses of forests, some
1
Author for correspondence; e-mail: mdelaestrella@gmail.com.
Manuscript received January 2014; revised manuscript received May 2014; elec-
tronically published August xx, 2014.
dominated by a single tree species (e.g., Gilbertiodendron dew-
evrei monodominant forest; Corlett and Primack 2011). About
half of all caesalpinioid genera (82 of 171) belong to the mono-
phyletic tribe Detarieae (Bruneau et al. 2008), which is pan-
tropical in distribution, but the majority of the genera are
confined to Africa and Madagascar (Mackinder 2005). One
consistently reported group within the Detarieae is the African
Berlinia clade (Bruneau et al. 2008), which is made up of 17
genera of medium to large trees. This clade includes a weakly
supported group made up of Didelotia,Plagiosiphon,Librev-
illea, and Gilbertiodendron, sister to a large clade that includes
Anthonotha,Englerodendron,Oddoniodendron (recently re-
viewed by Breteler 2006, 2008, 2010, 2011), Berlinia,Isob-
erlinia,Microberlinia (Mackinder and Pennington 2011), and
the ‘‘Babijt’’ group; the latter, as delimited by Wieringa and
Gervais (2003), includes Brachystegia,Aphanocalyx,Bikinia,
Julbernardia,Icuria, and Tetraberlinia. Many of the tree spe-
cies in the African forests belong to the Berlinia clade, and this
is the most important group of trees in the lowland evergreen
rainforest (Wieringa 1999) and a dominant component of Af-
rican tropical forests (White 1983).
Gilbertiodendron is an endemic African genus with 30 spe-
cies and a dominant component in many African forests. All
q1
2INTERNATIONAL JOURNAL OF PLANT SCIENCES
Thursday Aug 28 2014 10:26 AM/IJPS42308/2014/175/8/jharrell/islesskitain/ritterd//ms review complete/1002/use-graphics/narrow/
default/
species of Gilbertiodendron are trees of primary rain forest
and gallery forest on well-drained or periodically inundated
soil (Le´onard 1957; Mackinder 2005; Estrella et al. 2012a).
The origin of the genus dates from J. Le´onard’s work in the
1950s. Le´onard transferred all African species previously de-
scribed within Macrolobium to four separate genera, three of
which he described as new—Gilbertiodendron (Le´onard
1952), Paramacrolobium (Le´onard 1954), and Pellegriniod-
endron (Le´onard 1955); the fourth was made up of the rein-
stated genus Anthonotha P. Beauv. (Le´onard 1955). Pellegri-
niodendron (one species) was transferred to Gilbertiodendron
following recently published phylogenetic analyses (Bruneau
et al. 2008) and a morphological reevaluation of its status
(Estrella et al. 2012a). A complete taxonomic history of the
genus can be found in Estrella and Devesa (2014). Species
belonging to Gilbertiodendron have leaflets with marginal or
submarginal glands, a pair of bracteoles that encloses the
flower bud, five sepals, and five petals. The lateral and abaxial
petal pairs are reduced, alternate to the sepals and similar to
them. There is a single well-developed adaxial petal, with an
unguiculate base and bilobed apex. The androecium is usually
formed of six staminodes and three stamens fused at the base
on a short, fleshy tube, but there are a few exceptions, such
as G. splendidum, which has nine well-developed stamens. The
ovary, located on a short stipe, develops into a pod with 1–4
longitudinal nerves, the valves twisting when mature to dis-
perse the seeds explosively (Cowan and Polhill 1981; Estrella
et al. 2012a), similar to the ovary described for Tetraberlinia
(van der Burgt 1997).
Gilbertiodendron is a genus that is important both econom-
ically (Burkill 1995), with species used for timber or traditional
medicine, and ecologically, with species growing gregariously,
forming large forest stands (Poorter et al. 2004; Estrella et al.
2012c; van der Burgt et al. 2012; Estrella and Devesa 2014).
The highest concentration in both number of species and mor-
phological variation is found in the Guineo-Congolian region,
particularly in Gabon, which is a high-diversity area for le-
gumes (Sosef et al. 2006; Estrella et al. 2012b). At least 18
species of Gilbertiodendron have been reported in this area,
with 10 species considered endemic or near endemic (i.e., spe-
cies for which it was estimated that 85% of the distribution
area falls within Gabon; Sosef et al. 2006). Several additional
species new to science are to be described during the ongoing
taxonomic revision of the genus (van der Burgt et al., forth-
coming).
Although Gilbertiodendron plays a central role in many Af-
rican forest areas, as noted by Le´onard (1957), the advance-
ment of taxonomic knowledge and species delimitation within
the genus has been hampered by a lack of adequate material
for study, especially fruiting specimens. Recently, new collec-
tions have become available, resulting in, for example, the
recognition of many new species (Estrella et al. 2012c; van der
Burgt et al. 2012, forthcoming; Estrella and Devesa 2014), but
even now there is a lack of good material for many taxa,
preventing adequate descriptions (Estrella and Devesa 2014).
With the addition of recent collections, the objectives of the
current study are to complement and pursue the ongoing effort
of the past few years on the taxonomy of this genus (Estrella
et al. 2012a, 2012c; van der Burgt et al. 2012; Estrella and
Devesa 2014) with a densely sampled species-level phyloge-
netic analysis in order to identify species groups within Gil-
bertiodendron and to resolve relationships among genera con-
sidered closely related to Gilbertiodendron (Bruneau et al.
2008). More specifically, the aims of the present study are (1)
to test the monophyly of Gilbertiodendron as currently cir-
cumscribed, in particular to determine whether the inclusion
of Pellegriniodendron within Gilbertiodendron based on mor-
phological characters is supported by the phylogenetic results,
and (2) to assess the phylogenetic relationships of Gilbertio-
dendron with the other genera of the Berlinia clade. To address
these issues, we sequenced and analyzed the plastid trnL intron
and trnL-F spacer and the nuclear ribosomal internal tran-
scribed spacers (ITS) in Gilbertiodendron and related genera.
Material and Methods
Taxon Sampling
A total of 85 accessions, representing 28 species of Gil-
bertiodendron (including 10 accessions of G. diphyllum), 9
accessions representing 4 of the 5 recognized Plagiosiphon spe-
cies, 7 accessions representing 4 of 11 Didelotia species, and
1 accession of Librevillea klainei (monotypic genus) were sam-
pled. This is the widest sampling of Gilbertiodendron,Pla-
giosiphon, and Didelotia assembled to date for phylogenetic
analysis (the appendix provides voucher information). To test
the monophyly of Gilbertiodendron and examine relationships
among Detarieae genera, particularly in the Berlinia clade, we
sampled 10 other Berlinia clade genera and 5 representative
genera of the Brownea clade (sensu Bruneau et al. 2008), in-
cluding Macrolobium, which in the past has been considered
a close relative of Gilbertiodendron (see appendix for refer-
ences). Barnebydendron riedelii (Tul.) J. H. Kirkbr. and Schotia
latifolia Jacq. were included as outgroup taxa to root the trees
(Bruneau et al. 2008). Samples collected in the field were pre-
served in silica gel, and other samples were obtained from dried
herbarium specimens.
Molecular Methods
DNA extraction of herbarium and silica gel–dried material
was done using a modified protocol from Ky et al. (2000)
rescaled for a total 3 mL of nucleic extraction buffer (15 mM
Tris, 2 mM EDTA, 80 mm KCl, 20 mM NaCl, 2% b-mer-
captoethanol, PPVP 2%, 0.5% Trixon-X100), and the pellet
was re-covered in 2 mL of lysis buffer pH 8 (0.1 M Tris, 0.02
M EDTA, 1.25 M NaCl, MATAB 4%).
The polymerase chain reaction (PCR) amplification mix in
reaction volumes of 50 mL contained four units of Taq DNA
polymerase, 1#Taq DNA polymerase buffer with 1.5 mmol
MgCl
2
(New England Biolabs, Pickering, Ontario, Canada),
200 mmol/L of each dNTP (Fermentas, Burlington, Ontario,
Canada), 3 mmol/L of each primer, and 50–100 ng of genomic
DNA. For samples that were difficult to amplify, BSA (0.1 mg/
mL; New England BioLabs, Ipswich, MA), Tween 20 (0.03%;
J.-T. Baker, Phillipsburg, NJ), and pure DMSO (4%; Fisher
Scientific, Ottawa, Ontario, Canada) were also added to the
mix.
To maximize the yield of PCR products for the trnL and
trnL-F regions, a nested PCR method was used, as described
by Sinou et al. (2009), with the primer pairs “c”-“f,” and then
q2
DE LA ESTRELLA ET AL.—GILBERTIODENDRON PHYLOGENY (LEGUMINOSAE) 3
Thursday Aug 28 2014 10:26 AM/IJPS42308/2014/175/8/jharrell/islesskitain/ritterd//ms review complete/1002/use-graphics/narrow/
default/
Table 1
Sequence Characteristics, Parsimony and Bayesian Analysis Statistics, and Choice of Evolutionary Model for Each of the Matrices and
Subsets of the Matrices Analyzed in the Phylogenetic Analysis of the Genus Gilbertiodendron and Related Genera
ITS trnL intron trnL-F spacer Combined
No. sequences 127 110 92 127
Aligned length (bp) 1084 966 629 2679
Indels 286 68 72 426
Excluded characters (%) 71 (6.6) 106 (11) 0 177 (6.6)
Variable characters (%) 825 (63.5) 242 (26.7) 217 (31) 1284 (44.2)
Parsimony informative characters (%) 450 (34.6) 99 (10.9) 76 (10.8) 619 (21.3)
CI, RI, length .50, .75, 2227 .82, .90, 326 .78, .85, 303 .55, .75 2955
% GC content 44.5 20.9 27.2 31.9
Evolutionary model, AIC GTRIGGTRIGGTRIGGTRIG
Note. CI pconfidence index, RI pretention index, AIC pAkaike Information Criterion.
the plastid trnL (UAA) intron and the spacer between trnL
(UAA) and trnF (GAA) were, respectively, amplified and se-
quenced with the primer pairs “c”-“d” and “e”-“f,” as de-
scribed in Taberlet et al. (1991). For the ITS region, amplifi-
cations were performed with the “AB101” and “AB102”
primers (Sun et al. 1994; Douzery et al. 1999). Conditions for
the amplification of the trnL-F region were as follows: 5 min
of initial denaturation at 95C, followed by 30 cycles of 30 s
at 95C, 45 s at 50C, and 90 s at 72C, with a final step of
7 min at 72C. For the ITS region, PCR amplification condi-
tions were 4 min of initial denaturation at 94C, followed by
32 cycles of 30 s at 94C, 45 s at 53C, and 90 s at 72C,
with a final step of 7 min at 72C.
Sequencing was performed with Big Dye Terminator 3.1
chemistry on an ABI 3730xl DNA Analyzer (Applied Biosys-
tems, Carlsbad, CA) at the Genome Quebec facilities (Mon-
treal, Canada). Sequences were assembled and edited with Ge-
neious 4.8.5 (Biomatters, http://www.geneious.com).
Phylogenetic Analyses
Sequence alignment was performed with Geneious 4.8.5
(Biomatters, http://www.geneious.com) using the default pa-
rameters. Alignments were then verified and modified manu-
ally where inconsistencies were found. Nonautapomorphic
simple gaps (indels) were scored as separate presence/absence
characters, following Simmons and Ochoterena (2000), as im-
plemented in SeqState 1.4.1 (Mu¨ller 2005).
Individual matrices (plastid and nuclear markers) were an-
alyzed separately for exploratory purposes, and a concatenated
data matrix (cpnuc) was analyzed to generate the phyloge-
netic tree. The trnL intron matrix included sequences from
110 accessions (19 genera), the trnL-F intergenic spacer matrix
included data from 92 accessions (6 genera), and the ITS ma-
trix included 127 accessions (21 genera).
Parsimony analyses were performed with PAUP*, version
4.0b10 (Swofford 2003). A first heuristic search was per-
formed with 1000 replicates of random addition sequence, tree
bisection-reconnection (TBR) branch-swapping, retaining only
five most parsimonious trees at each replicate. Starting with
the trees kept in memory from this initial analysis, a second
heuristic search was performed with TBR and a limit of
100,000 trees saved. Because this second analysis uses the to-
pologies obtained initially, it permits the investigation of more
optimal topologies than a “one-step” analysis (e.g., Davis et
al. 2004). Branch support was estimated from 5000 bootstrap
replicates under a heuristic strategy with one random addition-
sequence replicate, TBR branch-swapping, and a maximum
number of trees set at 100.
The Bayesian analyses were implemented in a parallel ver-
sion of Mr. Bayes (ver. 3.2.1; Ronquist and Huelsenbeck 2003).
Computations were made on the supercomputer Cottos from
the Universite´ de Montre´ al, managed by Calcul Que´ bec and
Compute Canada. The GTRIGnucleotide substitution
model was selected using the Akaike Information Criterion
(Akaike 1974) as implemented in ModelTest (ver. 3.7; Posada
and Crandall 1998) for all DNA nucleotide partitions. The
model for the gap partition was set to “restriction data” (F81-
like model), as suggested by Ronquist and Huelsenbeck (2003).
The Bayesian estimation consisted of two independent runs,
each for 50 #10
6
generations, sampling trees and parameters
every thousandth generation. Each run consisted of four si-
multaneous Monte Carlo Markov chains and four swaps per
generation. All sample points prior to reaching stationarity of
the chains were discarded (equivalent to discarding the first
5#10
6
generations as burn-in). Convergence was assessed
by comparing majority-rule consensus trees from the two anal-
yses and using Tracer (ver. 1.5; Rambaut and Drummond
2007) to compare density plots of the estimated parameters
and of the likelihoods from the two analyses. The posterior
probabilities for individual clades were compared for congru-
ence and summarized on a majority-rule consensus.
Results
Sequence Characteristics
Length, number of indels, number of variable characters,
and statistics for the phylogenetic analyses are given in table
1 for each of the regions studied (individual and combined
matrices). The trnL intron was 470–700 bp in length, the trnL-
trnF spacer was 450–600 bp, and the ITS region was gen-
erally 780–930 bp in length. A total of 619 (21.3%) parsimony
informative characters were obtained from the combined anal-
ysis (table 1), with a considerably higher proportion from the
ITS region (34.6%) than from the trnL intron (10.9%) or the
trnL-trnF spacer (10.8%).
Thursday Aug 28 2014 10:26 AM/IJPS42308/2014/175/8/jharrell/islesskitain/ritterd//ms review complete/1002/use-graphics/nar-
row/default/
Fig. 1 Phylogenetic analysis of the chloroplast trnL intron and trnL-trnF intergenic spacer and the nuclear internal transcribed spacers for Gilbertiodendron and related caesalpinioid genera.
Bayesian majority-rule consensus derived from 45,000 trees kept after reaching stationarity in two independent analyses. Posterior probabilities (mostly above branches) and bootstrap support values
from the parsimony analysis (mostly below branches) are noted. Branches in bold are those supported by a 1.0 posterior probability; clades that are unresolved in the parsimony analysis are indicated
by broken lines. For species represented by two specimens, the collector name and the collection number are indicated after the species name.
DE LA ESTRELLA ET AL.—GILBERTIODENDRON PHYLOGENY (LEGUMINOSAE) 5
Thursday Aug 28 2014 10:26 AM/IJPS42308/2014/175/8/jharrell/islesskitain/ritterd//ms review complete/1002/use-graphics/narrow/
default/
(Fig. 1, continued)
Phylogenetic Analysis
In both the parsimony and Bayesian analyses, the consensus
trees resulting from the analysis of the ITS data alone (not
shown) were better resolved than the consensus trees from the
trnL and trnL-trnF analyses alone (not shown), but the three
analyses yielded topologies with equivalent relationships
among groups. The parsimony analysis of the combined data
set reached the maximum number of trees retained in memory
(length p2955, confidence index p0.55, retention index p
0.75) and yielded a poorly resolved strict consensus tree (see
fig. 1). The Bayesian majority-rule consensus tree of the com-
bined matrix yielded a topology among groups similar to that
obtained from the parsimony analysis of the combined data
but with better resolution (fig. 1, with posterior probabilities
indicated above the branches).
Our analysis places the genera Microberlinia and Oddon-
iodendron (fig. 1a, bootstrap support [BS] p72%, posterior
probability [PP] p1.00) as a sister clade to the other Berlinia
clade genera. The monophyletic Babijt group defined by Wier-
inga and Gervais (2003) is weakly supported as sister to the
q3
q4
6INTERNATIONAL JOURNAL OF PLANT SCIENCES
Thursday Aug 28 2014 10:26 AM/IJPS42308/2014/175/8/jharrell/islesskitain/ritterd//ms review complete/1002/use-graphics/narrow/
default/
Didelotia species included in our analyses (BS !50%, PP p
0.74). A clade including all accessions of Berlinia,Isoberlinia,
and Anthonotha is placed as sister to Librevillea klainei (BS
!50%, PP p0.99), and this entire clade is moderately sup-
ported (BS p61%, PP p1) as sister to a group that includes
the American Brownea clade sister to the Plagiosiphon species
sampled (BS p60%, PP p0.84) and a clade that constitutes
all the Gilbertiodendron species.
In our analysis, Gilbertiodendron is strongly supported as
monophyletic (fig. 1b;BSp100%, PP p1.00) including all
accessions of G. diphyllum previously recognized as Pellegri-
niodendron diphyllum. Within Gilbertiodendron (fig. 1b), sev-
eral groups are resolved, but many with low support. The west
tropical African Gilbertiodendron species are resolved into two
clades, the “aylmeri” clade (BS p67%, PP p0.82) and the
“ivorense” clade (!50%, 0.69), both made up of species that
grow in the tropical forest from Guinea to east Ghana (with
the exception of G. demonstrans, which occurs in central Af-
rica). The “dewevrei” clade (BS !50%, PP p0.78), with
species from central Africa, includes G. dewevrei,G. imen-
oense,G. stipulaceum,G. grandistipulatum, and G. mayo-
mbense, as well as unidentified Gilbertiodendron samples. The
strongly supported “ogoouense complex” (BS p96%, PP p
1.00) includes a group of species identified as G. klainei,G.
ogoouense, and G. brachystegioides, along with several new
species to be described. All three samples of G. preussii group
in a moderately supported clade, but those of its close relative,
G. diphyllum, are not monophyletic (fig. 1b) in the combined
plastid and ITS analyses. However, in the ITS analyses alone
(results not shown), G. diphyllum samples form a single clade.
Discussion
Generic Relationships of Gilbertiodendron
In our analyses, Gilbertiodendron is resolved as a mono-
phyletic genus moderately supported as sister to a clade con-
stituted of Plagiosiphon and the American Brownea clade. Pre-
vious phylogenetic studies by Bruneau et al. (2008) placed
Gilbertiodendron in a poorly supported clade with three gen-
era—Didelotia, Librevillea, and Plagiosiphon—but with no
resolution among the four genera. Although we include these
3 genera, as well as 10 other Berlinia clade genera, our analyses
do not support a monophyletic Berlinia clade because of the
nested position of the Brownea clade. However, as in all other
phylogenetic analyses of Detarieae, generic-level relationships
in the Amherstieae clade are poorly resolved (LPWG 2013a).
Nevertheless, Plagiosiphon, a genus of five species of trees and
shrubs that grow gregariously in lowland forests and along
rivers (Mackinder 2005), is here supported as monophyletic
with four of the five species sampled (fig. 1a). Didelotia is also
supported as monophyletic (fig. 1a), but it occurs, albeit with
little support, as sister to the Babijt clade (fig. 1a)ofWieringa
and Gervais (2003). The monospecific genus Librevillea is here
resolved as sister to representatives of three other genera of
the Berlinia group (fig. 1a), a relationship that is supported
only in the Bayesian analysis. Thus, despite better species-level
sampling with more samples per species, relationships of Gil-
bertiodendron within the Amherstieae clade remain unclear.
Gilbertiodendron was included within tribe Amherstieae by
Le´onard (1957) based on the position of the bracteoles at the
upper part of the pedicel. Cowan and Polhill (1981) maintained
Gilbertiodendron within that tribe, but they questioned whether
Amherstieae was monophyletic. Breteler (1995) proposed a
modified classification that recognized Gilbertiodendron (in-
cluding Pellegriniodendron and 21 more genera) within the
newly delimited tribe Macrolobieae. Phylogenetic analysis of
trnL data confirmed the position of Gilbertiodendron inaMa-
crolobieae clade (Bruneau et al. 2001), which largely but not
completely corresponded to Breteler’s (1995) delimitation of
tribe Macrolobieae. Subsequently, this lineage made up of the
majority of the genera (but not Macrolobium) was renamed the
Berlinia clade (Bruneau et al. 2008). Mackinder (2005), in a
synopsis of all previous morphological and molecular studies,
considered both Detarieae and Amherstieae (including the Ber-
linia clade) within the tribe Detarieae sensu lato.
Species Relationships within Gilbertiodendron
Our analyses support Gilbertiodendron as monophyletic,
including all accessions of G. diphyllum previously recognized
as Pellegriniodendron. Several of the characters used by Le´-
onard (1957) and Cowan and Raven (1981) to distinguish
between Pellegriniodendron and Gilbertiodendron were found
to be of weak or no value (Estrella et al. 2012a). Le´onard
(1957) and Cowan and Polhill (1981) characterized Pellegri-
niodendron diphyllum (G. diphyllum) by the presence of sti-
pels on leaflets, but Estrella et al. (2012a) found several mature
specimens of G. unijugum with “stipels” present at the base
of the petiolules and in seedlings of other species. Gilbertio-
dendron diphyllum is an easily identifiable species with a single
normal pair of leaflets and a basal pair that is reduced to small
structures, sometimes referred to as stipels. The presence of
submarginal crateriform glands on the leaflets blades of G.
diphyllum is the only character that can be used to consider
Pellegriniodendron as a segregate genus. Extrafloral nectaries
are common in other genera of the Detarieae, and in some
groups, this feature is an informative taxonomic character for
species-level delimitations (e.g., in Daniellia; Estrella et al.
2010). In the Detarieae, these crateriform glands are usually
present on the leaflet blades, as reported in the Babijt clade
(fig. 1b)inAphanocalyx (Wieringa 1999), but in Gilbertio-
dendron, crateriform glands are found only in G. diphyllum,
which is sister to the Gilbertiodendron clade (fig. 1b). The
remaining Gilbertiodendron species have a different and pos-
sibly autapomorphic type of gland, located on the leaflet mar-
gins and not found in any other Detarieae genera. These mar-
ginal glands have been used as a key character for the easy
identification of the genus (e.g., Aubre´ville 1968). Although
the function and nature of the leaflet margin glands is not clear
(E. Smets, personal communication), their presence may be
related to ants, which have been reported as living in associ-
ation with Gilbertiodendron species (e.g., fig. 1 from Estrella
and Devesa 2014).
The species of Gilbertiodendron endemic to west tropical Af-
rica (Upper Guinea: Senegal to Togo) form two clades, but the
relationship between these two clades is not resolved (fig. 1b).
Clade aylmeri includes the recently published species G. ton-
kolili, which cannot be differentiated vegetatively from G. bil-
ineatum but is considered distinct from this species based on
q5
DE LA ESTRELLA ET AL.—GILBERTIODENDRON PHYLOGENY (LEGUMINOSAE) 7
Thursday Aug 28 2014 10:26 AM/IJPS42308/2014/175/8/jharrell/islesskitain/ritterd//ms review complete/1002/use-graphics/narrow/
default/
Fig. 2 Distribution of the genus Gilbertiodendron in Africa: 1 pwest tropical Africa, 2 pwest-central Africa, 3 pcentral Africa. The
arrow indicates the location of the Dahomey Gap biogeographic area.
floral and fruiting characters (Estrella et al. 2012c). Our analysis
resolves the two confirmed accessions of G. tonkolili as most
closely related to G. aylmeri, with strong support. In the sister
clade, G. bilineatum occurs with a recently described species,
G. jongkindii (Estrella and Devesa 2014), which is morpholog-
ically similar to G. obliquum, a species that is resolved as part
of the ivorense clade, the second clade of west tropical African
species (fig. 1b). Gilbertiodendron jongkindii has a symmetric
leaflet base and sepals that are densely hairy at the margins, two
characters that differentiate it from G. obliquum. In addition to
G. obliquum, the ivorense clade includes four other west tropical
African species—G. ivorense,G. limba,G. robynsianum, and
G. sp. nov. Jongkind 8884—and the central African G. de-
monstrans, but with a broad distribution that extends to Nigeria.
These phylogenetic results support the taxonomic framework
proposed for the western African species proposed by Estrella
and Devesa (2014), who recognize G. tonkolili and G. jong-
kindii as different species (fig. 1b).
All four accessions of G. unijugum form a strongly sup-
ported monophyletic group, weakly resolved as sister to one
of the three specimens of G. mayombense sampled (fig. 1b).
This G. mayombense specimen from the Korup National Park
(Cameroon) may represent a segregate species distinct from
the central African G. mayombense (Angola, Gabon, and D.
R. Congo), which group with the dewevrei clade. A detailed
comparative study of these samples will be done during the
ongoing taxonomic revision.
Of the 30 species within Gilbertiodendron, only G. di-
phyllum and G. preussii have a distribution that reaches both
margins of the Dahomey Gap (fig. 2). The Dahomey Gap is
a woodland and wooded grassland region from east Ghana to
Benin (Booth 1958) that results from a climatic anomaly as-
sociated with low precipitation (Salzmann and Hoelzmann
2005). Although the Dahomey Gap area is presently covered
with agricultural land, savannah, and deteriorated dry forest,
it has been suggested to be a barrier to rain forest species
dispersal (Hawthorne and Jongkind 2006), and within it, rem-
nant forest patches are considered biodiversity refugia that
should be conserved (Backe´us 1992; Chaı¨r et al. 2011). This
distribution on both sides of the Dahomey Gap is uncommon
for a Detarieae species (found in only 33 out of the 295 in-
digenous Detarieae species from west, west-central, and central
Africa; Lock 1989), probably as a consequence of the relatively
short and limited maximum dispersal distance of the ballistic
seeds (van der Burgt et al. 2012). Despite this unusual distri-
bution, no significant morphological differences were observed
between specimens from west tropical Africa (Senegal to
Ghana) and those from west-central Africa (Cameroon, Equa-
torial Guinea, and Gabon; Estrella et al. 2012a; Estrella and
Devesa 2014). Both G. diphyllum and G. preussii have been
reported from secondary forests, swampy areas, and near man-
groves, possibly explaining the wide distribution of these two
species beyond the Dahomey Gap, since small pockets of for-
ests along swamps and rivers have been far more widespread
during periods of forest contractions than the poper refuges
of dry-land rain forest. However, all four specimens of G.
diphyllum from west tropical Africa group into a single clade,
distinct from the specimens from west-central Africa (fig. 1b),
indicating the likelihood of a genetic isolation between the two
areas of the distribution of this species.
The dewevrei clade (fig. 1b) is a polytomy that includes
accessions representing at least five species from west-central
q6
8INTERNATIONAL JOURNAL OF PLANT SCIENCES
Thursday Aug 28 2014 10:26 AM/IJPS42308/2014/175/8/jharrell/islesskitain/ritterd//ms review complete/1002/use-graphics/narrow/
default/
and central Africa, with a broad morphological variability. The
dewevrei clade includes G. grandistipulatum,G. mayombense,
G. imenoense,G. stipulaceum, and G. dewevrei. Gilbertio-
dendron dewevrei forms extensive patches of monodominant
forest (Aubre´ville 1970; van Valkenburg et al. 1998; Peh et al.
2011) of great ecological and conservation importance (Peh et
al. 2011). In studies on monodominant forests of G. dewevrei
in central Africa, Peh et al. (2011) showed that dominance is
not related to soil type but rather to seed-dispersal capacity
or to the ectomycorrhizal status of the species. Other species
of Gilbertiodendron form codominant patches, where they
grow mixed with other species of Caesalpinioideae, as well as
with trees from other families. Van der Burgt et al. (2012)
hypothesized that the codominant patches are the consequence
of the relatively short and strictly limited maximum dispersal
distance of the ballistic seeds typical of most Detarieae, in-
cluding all Gilbertiodendron species.
In the ogoouense complex clade, the unclear position of
samples previously identified either as G. ogoouense,G. bra-
chystegioides,G. klainei,orG. newberyi is in accordance with
the result of the morphological study of the group suggesting
the existence of several new taxa. In ongoing taxonomic re-
visions (van der Burgt et al., forthcoming), taxa of the
ogoouense complex will be divided into several new species
(fig. 1b). A common characteristic of taxa in the complex is
the presence of small flowers and congested inflorescences. The
specimens within this group share the absence of a gland at
the apex of the bracteoles, in contrast to other Gilbertioden-
dron species where the apical glands within bracteoles can be
more than 3 mm in length (Estrella and Devesa 2014).
Ongoing agriculture and clear-cutting of forest leading to
severe habitat loss and degradation have been identified as
well-known threats to Gilbertiodendron populations, and
these continue to contribute to their decline. In order to be
considered for a conservation rank, a species should ideally
be described to reflect evolutionary entities, and the species-
level phylogeny presented here provides some of the back-
ground information needed to establish the adequate assess-
ments. Of the 30 species within the genus, only 8 have been
assessed so far under the International Union for Conservation
of Nature (IUCN) Red List criteria. From west tropical Africa
(fig. 2), five species are assessed: G. bilineatum,G. robyn-
sianum, and G. splendidum are classified as vulnerable (VU);
G. limba is considered near threatened; and the recently pub-
lished species G. tonkolili is assessed as critically endangered
(Estrella et al. 2012c; IUCN 2013). This would suggest that
the aylmeri and ivorense clades include the most endangered
groups within the genus, but this may reflect the bias in the
sampling assessment with a greater focus on the west tropical
African flora. For west-central Africa and central Africa (fig.
2), three species were assessed: G. klainei (VU), G. newberyi
(endangered), and G. pachyanthum (VU; van der Burgt et al.
2012; IUCN 2013). None of the other species of Gilbertio-
dendron have been assessed, and of the eight existing assess-
ments, five were made in 1998 and need reevaluation in light
of new data. The new data derived from this phylogenetic
study and the ongoing taxonomic treatment of the genus will
help establish the adequate assessments for the genus, which
includes many timber species currently overexploited in trop-
ical Africa.
Acknowledgments
We wish to thank the staff of the cited herbaria for their
support during our visits and for loan of material as well as
Annie Archambault, Marielle Babineau, Iva´n de la Providen-
cia, Fatima El Ayadi, Edeline Gagnon, Carole Sinou, Royce
Steeves, Amy Vandal, and Erin Zimmerman for their help at
the Institut de Recherche En Biologie Ve´ge´tale (IRBV) labs at
Montreal University. Manuel de la Estrella was funded by a
Juan de la Cierva grant (JCI-2009-05243) and visited the IRBV
with a Jose´ Castillejo grant (JC2011-0309), both from the
Spanish Ministry of Education; he visited the K, P, and WAG
herbaria under the FP7-funded Integrated Infrastructure Ini-
tiative SYNTHESYS (grants GB-TAF-1006, FR-TAF 2194,
and NL-TAF 2377, respectively). Work undertaken in this
study was also funded by a grant from the Natural Sciences
and Engineering Research Council of Canada to Anne Bruneau
and a travel award from the Que´bec Centre for Biodiversity
Science to de la Estrella. The operation of the supercomputer
Cottos is funded by the Canada Foundation for Innovation,
NanoQue´bec, Re´ seau de Me´ decine Ge´ne´tique Applique´e, and
the Fonds de Recherche du Que´bec–Nature et Technologies.
We would like to thank the two anonymous reviewers for their
valuable comments and the editor in chief, Patrick S. Heren-
deen, for his comments and help.
Appendix
Specimens Studied, Voucher Information, and Genbank Accession Numbers for the
Phylogenetic Analysis of the Genus Gilbertiodendron and Related Genera
All vouchers are deposited at the indicated herbaria. Sequences are compiled from previous works, indicated by superscript
lowercase letters: (a) Bruneau et al. (2001), (b) Gervais and Bruneau (2002), (c) Fouge`re-Danezan et al. (2007), (d) Redden et
al. (2010), and (e) Mackinder and Pennington (2011). Data are ordered by ITS, trnL intron, and trnL-F spacer. Missing data
are denoted with a dash.
Outgroup
Barnebydendron riedelii (Tul.) J.H. Kirkbr., Brammall s. n., no. 1953-35501(K), AY955777
c
, AF365209
a
, AY958491
c
;Schotia
latifolia Jacq., Bruneau s.n., no 1948-52201 (K), AY955775
c
, AF365124
a
, AY958528
c
.Related genera: Anthonotha macrophylla
P. Beauv., Wieringa 2996 (WAG), AF513653
b
, AF365234
a
,—;Aphanocalyx cynometroides Oliver, Wieringa 2355 (WAG),
DE LA ESTRELLA ET AL.—GILBERTIODENDRON PHYLOGENY (LEGUMINOSAE) 9
Thursday Aug 28 2014 10:26 AM/IJPS42308/2014/175/8/jharrell/islesskitain/ritterd//ms review complete/1002/use-graphics/narrow/
default/
AF513654
b
, AF365244
a
,—;Aphanocalyx microphyllus subsp. compactus (Hutchinson ex Lane-Poole) Wieringa, Breteler 13356
(WAG), AF513662
b
, AF365246
a
,—;Berlinia auriculata Benth., Wieringa 5283 (WAG), HM041834
e
,—,—;Berlinia congolensis
(Baker f.) Keay, Harris 8585 (E), HM041826
e
,—,—;Berlinia grandiflora (Vahl) Hutch. & Dalziel, Harris 2895 (E), HM041821
e
,
—, —; Bikinia durandii (F. Halle´ & Normand) Wieringa, Wieringa 3021 (WAG), AF513676
b
, AY116896
b
,—;Bikinia grisea
Wieringa, Breteler 13334 (WAG), AF513678
b
, AY116897
b
,—;Brachystegia longifolia Benth., Herendeen 21-xii-97-2 (US),
AF513687
b
,—,—;Elizabetha macrostachya Benth., Redden 3714 (US), FJ817516
d
, FJ817559
d
,—;Elizabetha speciosa Ducke,
Rodrigues & Coelha 4850 (US), FJ817521
d
,—,—;Heterostemon impar Spruce ex. Benth., Amaral et al. 229 (K), FJ817524
d
,
FJ817562
d
,—;Heterostemon mazarunensis Sandwith, Redden 3203 (US), FJ817526
d
,—,—;Isoberlinia doka Craib & Stapf,
Jongkind 2552a (WAG), AF513691
b
, AF365220
a
,—;Julbernardia brieyi (De Wild.) Troupin, Wieringa 3348 (WAG), AF513692
b
,
AF365264
a
,—;Macrolobium campestre Huber, Redden 3649 (US), FJ817499
d
, FJ817551
d
,—;Macrolobium gracile Spruce ex.
Benth., Redden 3687 (US), FJ817500
d
, FJ817552
d
,—;Macrolobium multijugum var. multijugum (DC.) Benth., Redden 3700
(US), FJ817502
d
, FJ817554
d
,—;Microberlinia brazzavillensis A. Chev., Wieringa 2516 (WAG), AF513697
b
, AF365222
a
,—;
Oddoniodendron normandii Aubrev., Breteler 12608 (WAG), AF513698
b
, AF365224
a
,—;Paloue riparia Pulle, Redden 1161A
(US), FJ817546
d
, EU361826
d
,—;Paloveopsis emarginata R. S. Cowan, Cid Ferreira et al. 916 (NY), FJ817533
d
, FJ817571
d
,
—; Tetraberlinia bifoliolata (Harms) Hauman, Wieringa 3304 (WAG), AF513700
b
, AF365227
a
,—.Study group: Didelotia
africana Baill., Sosef 2509 (WAG), KJ777183, KJ777375, KJ777285; Didelotia africana Baill., Wieringa 3549 (WAG), KJ777184,
KJ777376, KJ777286; Didelotia letouzeyii Pellegr., Wieringa 5065 (WAG), KJ777185, KJ777377, KJ777287; Didelotia letouzeyii
Pellegr., Wieringa 6120 (WAG), KJ777186, KJ777378, KJ777288; Didelotia letouzeyii Pellegr., Wieringa 6187 (WAG), KJ777187,
KJ777379, KJ777289; Didelotia minutiflora (A. Chev.) J. Le´onard, Wieringa 4819 (WAG), KJ777188, KJ777380, KJ777290;
Didelotia unifoliolata J. Le´onard, Sosef 2517 (WAG), KJ777189, KJ777381, KJ777291; Gilbertiodendron aylmeri (Hutch. &
Dalziel) J. Le´onard, Jongkind 7003 (WAG), KJ777191, KJ777383, KJ777293; Gilbertiodendron aylmeri (Hutch. & Dalziel) J.
Le´onard, van der Burgt 1458 (WAG), KJ777190, KJ777382, KJ777292; Gilbertiodendron bilineatum (Hutch. & Dalziel) J.
Le´onard, Jongkind 5437 (WAG), KJ777193, —, —; Gilbertiodendron bilineatum-tonkolili, Jongkind 8775 (WAG), KJ777192,
KJ777384, KJ777294; Gilbertiodendron brachystegioides (Harms) J. Le´onard, Breteler 14983 (WAG), KJ777194, KJ777385,
—; Gilbertiodendron brachystegioides (Harms) J. Le´onard, van der Burgt 801 (WAG), KJ777195, KJ777386, KJ777295; Gil-
bertiodendron brachystegioides (Harms) J. Le´onard, Wieringa 2242 (WAG), KJ777196, KJ777387, KJ777296; Gilbertiodendron
brachystegioides (Harms) J. Le´onard, Wieringa 6226 (WAG), KJ777197, KJ777388, KJ777297; Gilbertiodendron comb. nov.,
M’Boungou 144 (WAG), KJ777199, KJ777390, KJ777299; Gilbertiodendron comb. nov., M’Boungou 387 (WAG), KJ777200,
KJ777391, —; Gilbertiodendron comb. nov., Wieringa 6070 (WAG), KJ777201, —, —; Gilbertiodendron demonstrans (Baill.)
J. Le´onard, van der Burgt 731 (WAG), KJ777202, KJ777392, KJ777300; Gilbertiodendron demonstrans (Baill.) J. Le´onard,
Wieringa 2202 (WAG), KJ777203, —, KJ777301; Gilbertiodendron demonstrans (Baill.) J. Le´onard, Wieringa 2362 (WAG),
KJ777204, —, —; Gilbertiodendron dewevrei (De Wild.) J. Le´onard, Andel 4045 (WAG), KJ777205, KJ777393, KJ777302;
Gilbertiodendron dewevrei (De Wild.) J. Le´onard, van der Burgt 783 (WAG), KJ777206, KJ777394, KJ777303; Gilbertiodendron
dewevrei (De Wild.) J. Le´onard, van der Burgt 905 (WAG), KJ777207, KJ777395, KJ777304; Gilbertiodendron dewevrei (De
Wild.) J. Le´onard, Wieringa 3552 (WAG), KJ777208, KJ777396, KJ777305; Gilbertiodendron dewevrei (De Wild.) J. Le´onard,
Wieringa 6091 (WAG), KJ777209, KJ777397, KJ777306; Gilbertiodendron dewevrei (De Wild.) J. Le´onard, Wieringa 6261
(WAG), KJ777210, KJ777398, KJ777307; Gilbertiodendron diphyllum (Harms) Estrella & Devesa, Andel 3502 (WAG),
KJ777211, KJ777399, KJ777308; Gilbertiodendron diphyllum (Harms) Estrella & Devesa, Andel 4111 (WAG), KJ777212,
KJ777400, KJ777309; Gilbertiodendron diphyllum (Harms) Estrella & Devesa, Breteler 13402 (WAG), KJ777213, KJ777401,
KJ777310; Gilbertiodendron diphyllum (Harms) Estrella & Devesa, Breteler 13734 (WAG), KJ777214, KJ777402, KJ777311;
Gilbertiodendron diphyllum (Harms) Estrella & Devesa, Jongkind 9311 (WAG), KJ777215, KJ777403, KJ777312; Gilbertio-
dendron diphyllum (Harms) Estrella & Devesa, Louis 2481 (WAG), KJ777216, KJ777404, KJ777313; Gilbertiodendron di-
phyllum (Harms) Estrella & Devesa, Parren 306 (WAG), KJ777217, —, KJ777314; Gilbertiodendron diphyllum (Harms) Estrella
& Devesa, Wieringa 4105 (WAG), KJ777219, KJ777406, KJ777316; Gilbertiodendron diphyllum (Harms) Estrella & Devesa,
Wieringa 6218 (WAG), KJ777220, KJ777407, KJ777317; Gilbertiodendron diphyllum sp. nov., Breteler 10782 (WAG),
KJ777218, KJ777405, KJ777315; Gilbertiodendron grandistipulatum (De Wild.) J. Le´onard, Breteler 10428 (WAG), KJ777221,
KJ777408, KJ777318; Gilbertiodendron grandistipulatum (De Wild.) J. Le´onard, White 1118 (WAG), KJ777222, KJ777409,
KJ777319; Gilbertiodendron grandistipulatum (De Wild.) J. Le
´onard, Wieringa 4687 (WAG), KJ777223, KJ777410, KJ777320;
Gilbertiodendron imenoense (Pellegr.) J. Le´onard, Breteler 9974 (WAG), KJ777224, KJ777411, KJ777321; Gilbertiodendron
ivorense (A. Chev.) J. Le´onard, Jongkind 9210 (WAG), KJ777225, KJ777412, KJ777322; Gilbertiodendron ivorense (A. Chev.)
J. Le´onard, Stoop 19 (WAG), KJ777226, KJ777413, KJ777323; Gilbertiodendron jongkindii Estrella & Devesa, Jongkind 4502
(WAG), KJ777227, KJ777414, KJ777324; Gilbertiodendron jongkindii Estrella & Devesa, Jongkind 9350 (WAG), KJ777228,
KJ777415, KJ777325; Gilbertiodendron klainei (Pierre ex Pellegr.) J. Le´onard, Wieringa 1239 (WAG), KJ777229, KJ777416,
KJ777326; Gilbertiodendron klainei (Pierre ex Pellegr.) J. Le´onard, Wieringa 2443 (WAG), KJ777230, KJ777417, KJ777327;
Gilbertiodendron limba (Scott-Elliot) J. Le´onard, Jongkind 1438 (WAG), KJ777231, KJ777418, KJ777328; Gilbertiodendron
mayombense (Pellegr.) J. Le´onard, Breteler 14147 (WAG), KJ777232, KJ777419, KJ777329; Gilbertiodendron mayombense
(Pellegr.) J. Le´onard, Wieringa 1475 (WAG), KJ777233, —, —; Gilbertiodendron cf. mayombense, van der Burgt 595 (WAG),
KJ777198, KJ777389, KJ777298; Gilbertiodendron newberyi Burgt, van der Burgt 774 (WAG), KJ777234, KJ777420,
KJ777330; Gilbertiodendron newberyi Burgt, van der Burgt 776 (WAG), KJ777235, KJ777421, KJ777331; Gilbertiodendron
obliquum (Stapf) J. Le´onard, Jongkind 9972 (WAG), KJ777236, KJ777422, KJ777332; Gilbertiodendron ogoouense (Pellegr.)
10 INTERNATIONAL JOURNAL OF PLANT SCIENCES
Thursday Aug 28 2014 10:26 AM/IJPS42308/2014/175/8/jharrell/islesskitain/ritterd//ms review complete/1002/use-graphics/narrow/
default/
J. Le´onard, Breteler 10258 (WAG), KJ777237, KJ777423, KJ777333; Gilbertiodendron ogoouense (Pellegr.) J. Le´onard, Breteler
11450 (WAG), KJ777238, —, —; Gilbertiodendron ogoouense (Pellegr.) J. Le´onard, Breteler 11524 (WAG), KJ777239,
KJ777424, KJ777334; Gilbertiodendron ogoouense (Pellegr.) J. Le´onard, Schoenmaker 70 (WAG), KJ777242, KJ777427,
KJ777336; Gilbertiodendron ogoouense (Pellegr.) J. Le´onard, van der Burgt 775 (WAG), KJ777240, KJ777425, KJ777335;
Gilbertiodendron ogoouense (Pellegr.) J. Le´onard, Wieringa 1316 (WAG), KJ777243, —, —; Gilbertiodendron ogoouense
(Pellegr.) J. Le´onard, Wieringa 2777 (WAG), KJ777244, KJ777428, KJ777337; Gilbertiodendron ogoouense (Pellegr.) J. Le´onard,
Mpondzou 44 (WAG), KJ777241, KJ777426, —; Gilbertiodendron preussii (Harms) J. Le´onard, Andel 4118 (WAG), KJ777245,
KJ777429, KJ777338; Gilbertiodendron preussii (Harms) J. Le´onard, van der Burgt 794 (WAG), KJ777246, KJ777430,
KJ777339; Gilbertiodendron preussii (Harms) J. Le´onard, Wieringa 1600 (WAG), KJ777247, KJ777431, KJ777340; Gilberti-
odendron robynsianum Aubre´v. & Pellegr., Jongkind 4525 (WAG), KJ777248, KJ777432, KJ777341; Gilbertiodendron robyn-
sianum Aubre´v. & Pellegr., Jongkind 4942 (WAG), KJ777249, KJ777433, KJ777342; Gilbertiodendron robynsianum Aubre´v.
& Pellegr., Jongkind 5644 (WAG), KJ777250, KJ777434, KJ777343; Gilbertiodendron sp., Sosef 2502 (WAG), KJ777261,
KJ777445, KJ777353; Gilbertiodendron sp., Wieringa 1315 (WAG), KJ777262, —, —; Gilbertiodendron sp., Wieringa 3793
(WAG), KJ777263, KJ777446, KJ777354; Gilbertiodendron sp., Wieringa 5253 (WAG), KJ777264, KJ777447, KJ777355;
Gilbertiodendron sp. nov., Jongkind 8884 (WAG (Liberia), KJ777251, KJ777435, KJ777344; Gilbertiodendron sp. nov. A,
Bongou 105 (WAG), KJ777254, KJ777438, —; Gilbertiodendron sp. nov. A, Boungou 415 (WAG), KJ777255, KJ777439,
KJ777347; Gilbertiodendron sp. nov. A, Valkenburg 2635 (WAG), KJ777256, KJ777440, KJ777348; Gilbertiodendron sp. nov.
A, Valkenburg 2636 (WAG), KJ777257, KJ777441, KJ777349; Gilbertiodendron sp. nov. B, Breteler 14250 (WAG), KJ777258,
KJ777442, KJ777350; Gilbertiodendron sp. nov. B, Wieringa 6190 (WAG), KJ777259, KJ777443, KJ777351; Gilbertiodendron
sp. nov. B, Wieringa 6191 (WAG), KJ777260, KJ777444, KJ777352; Gilbertiodendron sp. nov.?, Wieringa 6099 (WAG),
KJ777253, KJ777437, KJ777346; Gilbertiodendron sp. nov., Wieringa 5057 (WAG), KJ777252, KJ777436, KJ777345; Gil-
bertiodendron splendidum (A. Chev. ex Hutch. & Dalziel) J. Le´onard, Jongkind 8782 (WAG), KJ777265, KJ777448, KJ777356;
Gilbertiodendron stipulaceum (Benth.) J. Le´onard, Schoenmaker 313 (WAG), KJ777266, KJ777449, KJ777357; Gilbertiodendron
stipulaceum (Benth.) J. Le´onard, Schoenmaker 385 (WAG), KJ777267, —, —; Gilbertiodendron stipulaceum (Benth.) J. Le´onard,
Wieringa 1644 (WAG), KJ777268, KJ777450, KJ777358; Gilbertiodendron tonkolili Burgt & Estrella, van der Burgt 1423
(WAG), KJ777269, KJ777451, KJ777359; Gilbertiodendron tonkolili Burgt & Estrella, van der Burgt 1457 (WAG), KJ777270,
KJ777452, KJ777360; Gilbertiodendron unijugum (Pellegr.) J. Le´onard, McPherson 15847 (WAG), KJ777272, KJ777454,
KJ777362; Gilbertiodendron unijugum (Pellegr.) J. Le´onard, van der Burgt 73 (WAG), KJ777271, KJ777453, KJ777361; Gil-
bertiodendron unijugum (Pellegr.) J. Le´onard, Wieringa 1075 (WAG), KJ777273, KJ777455, KJ777363; Gilbertiodendron un-
ijugum (Pellegr.) J. Le´onard, Wilde 10974 (WAG), KJ777274, KJ777456, KJ777364; Librevillea klainei (Pierre ex Harms) Hoyle,
Sosef 2505 (WAG), KJ777275, KJ777457, KJ777365; Plagiosiphon emarginatus (Hutch. & Dalziel) J. Le´onard, Wieringa 6067
(WAG), KJ777276, KJ777458, KJ777366; Plagiosiphon gabonensis J. Le´onard, Sosef 2503 (WAG), KJ777277, KJ777459,
KJ777367; Plagiosiphon gabonensis J. Le´onard, Wieringa 4400 (WAG), KJ777278, KJ777460, KJ777368; Plagiosiphon ga-
bonensis J. Le´onard, Wieringa 6137 (WAG), KJ777279, KJ777461, KJ777369; Plagiosiphon longitubus J. Le´onard, Wieringa
5864 (WAG), KJ777280, —, KJ777370; Plagiosiphon longitubus J. Le´onard, Wieringa 5870 (WAG), KJ777281, KJ777462,
KJ777371; Plagiosiphon multijugus J. Le´onard, Andel 4044 (WAG), KJ777282, KJ777463, KJ777372; Plagiosiphon multijugus
J. Le´onard, Wieringa 3813 (WAG), KJ777283, KJ777464, KJ777373; Plagiosiphon sp., Wieringa 4039 (WAG), KJ777284,
KJ777465, KJ777374.
Literature Cited
Akaike H 1974 A new look at the statistical model identification.
IEEE Trans Automat Contr 19:716–723.
Aubre´ ville A 1968 Le´gumineuses—Ce´salpinı¨oide´ es. Pages XX–XX in
Flore du Gabon. Vol 15. Muse´um National d’Histoire Naturelle, Paris.
——— 1970 Le´gumineuses—Ce´ salpinı¨oide´es. Pages XX–XX in Flore
du Cameroun. Vol 9. Muse´um National d’Histoire Naturelle, Paris.
Backe´us I 1992 Distribution and vegetation dynamics of humid sa-
vannas in Africa and Asia. J Veg Sci 3:345–356.
Booth AH 1958. The Niger, the Volta and the Dahomey Gap as geo-
graphic barriers. Evolution 12:48–62.
Breteler FJ 1995 The boundary between Amherstieae and Detarieae
(Caesalpinioideae). Pages 53–61 in M Crisp, JJ Doyle, eds. Advances
in legume systematics. Vol 7. Phylogeny. Royal BotanicGardens, Kew.
——— 2006 Novitates Gabonenses 56. Two Anthonotha species
from Gabon transferred to Englerodendron (Fabaceae, Caesalpi-
nioideae). Adansonia 28:106–111.
——— 2008 Anthonotha and Isomacrolobium (Leguminosae, Cae-
salpinioideae): two distinct genera. Syst Geogr Plants 78:137–144.
——— 2010 Revision of the African genus Anthonotha (Legumino-
sae, Caesalpinioideae). Plant Ecol Evol 143:70–99.
——— 2011 Revision of the African genus Isomacrolobium (Leg-
uminosae, Caesalpinioideae). Plant Ecol Evol 144:64–81.
Bruneau A, F Forest, PS Herendeen, BB Klitgaard, GP
Lewis 2001 Phylogenetic relationships in the Caesalpinioideae (Leg-
uminosae) as inferred from chloroplast trnL intron sequences. Syst
Bot 26:487–514.
Bruneau A, M Mercure, GP Lewis, P Herendeen 2008 Phylogenetic
patterns and diversification in the Caesalpinioid legumes. Botany
86:697–718.
Burkill HM 1995 The useful plants of west tropical Africa. Vol 3.
Families J–L. Royal Botanic Gardens, Kew.
Chaı¨r H, PO Duroy, P Cubry, B Sinsin, JL Pham 2011 Impact of past
climatic and recent anthropogenic factors on wild yam genetic di-
versity. Mol Ecol 20:1612–1623.
Corlett R, R Primack 2011 Tropical rain forests: an ecological and
biogeographical comparison. 2nd ed. Wiley, Oxford.
Cowan RS, RM Polhill 1981 Tribe 5. Amherstieae Benth. emend. J.
Le´on 1957. Pages 135–142 in RM Polhill, PH Raven, eds. Advances
in legume systematics. Vol 1. Royal Botanic Gardens, Kew.
Davis JI, DW Stevenson, G Petersen, O Seberg, LM Campbell, JV
q7
DE LA ESTRELLA ET AL.—GILBERTIODENDRON PHYLOGENY (LEGUMINOSAE) 11
Thursday Aug 28 2014 10:26 AM/IJPS42308/2014/175/8/jharrell/islesskitain/ritterd//ms review complete/1002/use-graphics/narrow/
default/
Freudenstein, DH Goldman, et al 2004 A phylogeny of the mono-
cots, as inferred from rbcL and atpA sequence variation, and a
comparison of methods for calculating jackknife and bootstrap val-
ues. Syst Bot 29:467–510.
Douzery EJP, AM Pridgeon, P Kores, HP Linder, H Kurzweil, MW
Chase 1999 Molecular phylogenetics of Diseae (Orchidaceae): a
contribution from nuclear ribosomal ITS sequences. Am J Bot 86:
887–899.
Estrella M, C Aedo, B Mackinder, M Velayos 2010 Taxonomic re-
vision of Daniellia (Leguminosae: Caesalpinioideae). Syst Bot 35:
296–324.
Estrella M, JA Devesa 2014 The genus GilbertiodendronJ. Le´ onard (Leg-
uminosae-Caesalpinioideae) in western Africa. Syst Bot 39:160–192.
Estrella M, JA Devesa, JJ Wieringa 2012aA morphological re-evalu-
ation of the taxonomic status of the genus Pellegriniodendron (Harms)
J. Le´ onard (Leguminosae-Caesalpinioideae-Detarieae) and its inclu-
sion in Gilbertiodendron J. Le´onard. S Afr J Bot 78:257–265.
Estrella M, RG Mateo, JJ Wieringa, B Mackinder, J Mun˜ oz 2012b
Legume diversity patterns in west central Africa: influence of species
biology on distribution models. PLoS ONE 7:e41526.
Estrella M, XM van der Burgt, BA Mackinder, JA Devesa, MS James, WD
Hawthorne 2012c Gilbertiodendron tonkolili sp. nov. (Leguminosae-
Caesalpinioideae) from Sierra Leone. Nord J Bot 30:136–143.
Fouge`re-Danezan M, S Maumont, A Bruneau 2007 Relationships
among resin-producing Detarieae s.l. (Leguminosae) as inferred by
molecular data. Syst Bot 32:748–761.
Gervais GYF, A Bruneau 2002 Phylogenetic analysis of a polyphyletic
African genus of Caesalpinioideae (Leguminosae): Monopetalanthus
Harms. Plant Syst Evol 235:19–34.
Hawthorne WD, CCH Jongkind 2006 Woody plants of western Af-
rican forests: a guide to the forest trees, shrubs and lianes from
Senegal to Ghana. Royal Botanic Gardens, Kew.
International Union for Conservation of Nature (IUCN) 2013 IUCN
red list of threatened species. Ver. 2013.1. http://www.iucnredlist.org.
Ky C-L, P Barre, M Lorieux, P Trouslot, S Akaffou, J Louarn, A
Charrier, S Hamon, M Noirot 2000 Interspecific genetic linkage
map, segregation distortion and genetic conversion in coffee (Coffea
sp.). Theor Appl Genet 101:669–676.
Lebrun J-P, AL Stork 1998 Analyse structurelle de la flore des An-
giosperms d’Afrique tropicale. Candollea 53:365–385.
Le´onard J 1952 Notulae systematicae. XII. Les genres Macrolobium
Schreb. et Gilbertiodendron J. Le´ onard au Congo Belge. Bull Jard
Bot Etat 22:185–191.
——— 1954 Notulae systematicae. XIV. Les genres Macrolobium
Schreb. et Gilbertiodendron J. Le´ onard en Afrique Tropicale. Bull
Jard Bot Etat 24:57–61.
——— 1955 Notulae systematicae. XVII. Les genres Anthonotha P.
Beauv. et Pellegriniodendron J. Le´onard en Afrique Tropicale (Cae-
salpinaceae). Bull Jard Bot Etat 25:201–203.
——— 1957 Genera des Cynometreae et des Amherstieae africaines
(Le´guminosae-Caesalpinioideae). Essai de blastogenie applique´a`la
syste´matique. Mem Acad R Belg Classe Sci 30:1–314.
Letouzey R 1968 Etude Phytoge´ ographique du Cameroun. Encyclo-
pe´ die Biologique 69. Lechevalier, Paris.
Lewis G, B Schrire, B Mackinder, M Lock 2005 Legumes of the
world. Royal Botanic Gardens, Kew.
Lock JM 1989 Legumes of Africa: a check-list. Royal Botanic Gar-
dens, Kew.
Legume Phylogeny Working Group (LPWG) 2013aLegume phylog-
eny and classification in the 21st century: progress, prospects and
lessons for other species-rich clades. Taxon 62:217–248.
——— 2013bTowards a new classification system for legumes: pro-
gress report from the 6th International Legume Conference. S Afr
J Bot 89:3–9.
Mackinder BA 2005 Detarieae sensu lato. Pages 69–109 in G Lewis,
B Schrire, B Mackinder, M Lock, eds. Legumes of the world. Royal
Botanic Gardens, Kew.
Mackinder BA, RT Pennington 2011 Monograph of Berlinia (Leg-
uminosae). Syst Bot Monogr 91:1–117.
Mu¨ller K 2005 SeqState: primer design and sequence statistics for
phylogenetic DNA data sets. Appl Bioinform 4:65–69.
Peh SB, J Lloyd, CA Quesada, SL Lewis 2011 Soil does not explain
monodominance in a central African tropical forest. PLoS ONE 6:
e16996.
Poorter L, F Bongers, FN Kouame´ , WD Hawthorne 2004 Biodiversity
of west African forests: an ecological atlas of woody plant species.
CABI, Oxon.
Posada D, KA Crandall 1998 ModelTest: testing the model of DNA
substitution. Bioinformatics 14:817–818.
Rambaut A, AJ Drummond 2007 Tracer. Ver. 1.4. http://beast.bio
.ed.ac.uk/Tracer.
Redden KM, PS Herendeen, KJ Wurdack, A Bruneau 2010
Phylogenetic relationships of the northeastern South American
Brownea clade of tribe Detarieae (Leguminosae: Caesalpinioideae)
based on morphology and molecular data. Syst Bot 35:524–533.
Ronquist F, JP Huelsenbeck 2003 MrBayes 3: Bayesian phylogenetic
inference under mixed models. Bioinformatics 19:1572–1574.
Salzmann U, P Hoelzmann 2005 The Dahomey Gap: an abrupt cli-
matically induced rain forest fragmentation in west Africa during
the late Holocene. Holocene 15:190–199.
Simmons MP, H Ochoterena 2000 Gaps as characters in sequence-
based phylogenetic analyses. Syst Biol 40:369–381.
Sinou C, F Forest, GP Lewis, A Bruneau 2009 The genus Bauhinia
s.l. (Leguminosae): a phylogeny based on the plastid trnL-trnF re-
gion. Botany 87:947–960.
Sosef MSM, JJ Wieringa, CCH Jongkind, G Achoundong, Y Azizet
Issembe, D Bedigian, RG Berg, et al 2006 Check-list des plantes
vasculaires du Gabon. Scr Bot Belg 35:1–438.
Sun Y, DZ Skinner, GH Liang, SH Hulbert 1994 Phylogenetic analysis
of Sorghum and related taxa using internal transcribed spacers of
nuclear ribosomal DNA. Theor Appl Genet 89:26–32.
Swofford DL 2003 PAUP*: phylogenetic analysis using parsimony
(*and other methods). Ver. 4.0b10. Sinauer, Sunderland, MA.
Taberlet P, L Gielly, G Pautou, J Bouvet 1991 Universal primers for
amplification of three non-coding regions of chloroplast DNA.Plant
Mol Biol 17:1105–1109.
van der Burgt XM 1997 Explosive seed dispersal of the rainforest
tree Tetraberlinia moreliana (Leguminosae—Caesalpinioideae) in
Gabon. J Trop Ecol 13:145–151.
van der Burgt XM, M Eyakwe, J Motoh 2012 Gilbertiodendron new-
beryi (Leguminosae: Caesalpinioideae), a new tree species from
Korup National Park, Cameroon. Kew Bull 67:1–7.
van der Burgt XM, BA Mackinder, JJ Wieringa, M de la Estrella
Forthcoming The Gilbertiodendron ogoouense species complex
(Leguminosae—Caesalpinioideae) in central Africa.
van Valkenburg JLCH, P Ketner, CM Wilks 1998 A floristic inventory
and preliminary vegetation classification of the mixed semi-ever-
green rain forest in the Minkebe region, North East Gabon. Adan-
sonia 20:139–162.
White F 1983 The vegetation of Africa. A descriptive memoir to ac-
company the Unesco/AETFAT/UNSO vegetation map of Africa.
Courvoisier, La Chaux-de-Fonds.
Wieringa JJ 1999 Monopetalanthus exit: a systematic study of Aphan-
ocalyx,Bikinia,Icuria,Michelsonia and Tetraberlinia (Leguminosae,
Caesalpinioideae). Wagening Agric Univ Pap 99–4:1–320.
Wieringa JJ, GYF Gervais 2003 Phylogenetic analyses of combined mor-
phological and molecular data sets on the Aphanocalyx-Bikinia-Tetra-
berlinia group (Leguminosae, Caesalpinioideae, Detarieae s.l.). Pages
181–196 in BB Klitgaard, A Bruneau, eds. Advances in legume system-
atics. Vol 10. Higher level systematics. Royal Botanic Gardens, Kew.
q8
12 INTERNATIONAL JOURNAL OF PLANT SCIENCES
Thursday Aug 28 2014 10:26 AM/IJPS42308/2014/175/8/jharrell/islesskitain/ritterd//ms review complete/1002/use-graphics/narrow/
default/
QUERIES TO THE AUTHOR
q1. I edited your article for clarity and agreement with
journal style, and minor revisions were made with-
out comment. Please review the text and advise
whether any of my changes are unacceptable. Any
layout problems or errors you see will be corrected
after the proof is returned.
q2. Should this be “80 mM KCl”?
q3. Here and in table 1, did spell out L, CI ,and RI
correctly? And, if Lplength, does the value require
a unit of measurement?
q4. If acceptable, we would like to add labels to figures
1a and 1b for ease of reading. They would appear
as labels a and b on the figures themselves. Is this
acceptable?
q5. Please add Cowan and Raven 1981 to the Liter-
ature Cited.
q6. Re. “the poper refuges”: is there a typo here?
q7. Please provide page numbers for Aubre´ville 1968
and 1970.
q8. Please provide updated publication information
for van der Burgt et al, forthcoming.
... 180 species that occur exclusively in Africa. Previous studies based on morphology and a few molecular markers have identified major lineages within this group, e.g. the "babjit" group (Wieringa, 1999), which also has been recovered in more recent phylogenetic analyses (de la Estrella et al., 2018(de la Estrella et al., , 2014. However, generic relationships have not been fully resolved and a limited number of species have been included for most genera within the Berlinia clade in previous studies (Bruneau et al., 2008;de la Estrella et al., 2014;Mackinder et al., 2013). ...
... Previous studies based on morphology and a few molecular markers have identified major lineages within this group, e.g. the "babjit" group (Wieringa, 1999), which also has been recovered in more recent phylogenetic analyses (de la Estrella et al., 2018(de la Estrella et al., , 2014. However, generic relationships have not been fully resolved and a limited number of species have been included for most genera within the Berlinia clade in previous studies (Bruneau et al., 2008;de la Estrella et al., 2014;Mackinder et al., 2013). Within the Berlinia clade, the genus Anthonotha and its two close relatives, Isomacrolobium and Englerodendron, are potentially the most problematic group that remains to be studied because several studies have D.I. Ojeda, et al. ...
... In order to represent the closest genera of our focal group, we also included eight of the 17 genera within the Berlinia clade. These eight genera have been previously identified as the closest related lineages of our focal group (de la Estrella et al., 2018(de la Estrella et al., , 2017(de la Estrella et al., , 2014. ...
Article
Full-text available
Detarioideae is well known for its high diversity of floral traits, including flower symmetry, number of organs, and petal size and morphology. This diversity has been characterized and studied at higher taxonomic levels, but limited analyses have been performed among closely related genera with contrasting floral traits due to the lack of fully resolved phylogenetic relationships. Here, we used four representative transcriptomes to develop an exome capture (target enrichment) bait for the entire subfamily and applied it to the Anthonotha clade using a complete data set (61 specimens) representing all extant floral diversity. Our phylogenetic analyses recovered congruent topologies using ML and Bayesian methods. Anthonotha was recovered as monophyletic contrary to the remaining three genera (Englerodendron, Isomacrolobium and Pseudomacrolobium), which form a monophyletic group sister to Anthonotha. We inferred a total of 35 transitions for the seven floral traits (pertaining to flower symmetry, petals, stamens and staminodes) that we analyzed, suggesting that at least 30% of the species in this group display transitions from the ancestral condition reconstructed for the Anthonotha clade. The main transitions were towards a reduction in the number of organs (petals, stamens and staminodes). Despite the high number of transitions, our analyses indicate that the seven characters are evolving independently in these lineages. Petal morphology is the most labile floral trait with a total of seven independent transitions in number and seven independent transitions to modification in petal types. The diverse petal morphology along the dorsoventral axis of symmetry within the flower is not associated with differences at the micromorphology of petal surface, suggesting that in this group all petals within the flower might possess the same petal identity at the molecular level. Our results provide a solid evolutionary framework for further detailed analyses of the molecular basis of petal identity.
... Samples collected in the field were preserved in silica gel, and other samples were obtained from dried herbarium specimens. We generated most of the sequences (including 475 sequences newly released for this study), and the sampling was completed with additional sequences produced by our research group in previous studies 30,36,39,40,44 which were downloaded from GenBank (http://www.ncbi.nlm.nih.gov/genbank/) to complete the taxon and gene sampling. To avoid the effects of missing data no sample was included that had fewer than two loci sequenced, and for this reason six genera that have been included in other studies (Brachycylix, Lebruniodendron, Micklewaitia, Michelsonia, Neoapaloxylon, Paloveopsis) are not included in our analyses. ...
... For the ITS/5.8 S region, amplifications were performed with the "AB101" and "AB102" primers 48,49 ; conditions for the amplification follow Estrella et al. 40 . The matK gene and the flanking 3′ intron region were amplified in one fragment using the primers trnK685F and trnK2Rdet 30 and the internal primers described in that study were used to sequence the most difficult samples. ...
... In the past several years a number of studies have been published that aim to understand relationships and evolution in this group (e.g. 1,[3][4][5]10,15,[28][29][30][31]33,[35][36][37][38][39][40][41]44,[72][73][74][75][76] ) and along with the new phylogenetic analysis presented here, we are in a position to present a formal tribal classification of Detarioideae that will provide the necessary framework to better understand the systematics and evolutionary origin of this lineage. Three genera, Schotia, Goniorrhachis and Barnebydendron, always appear among the early branching clades within Detarioideae [29][30][31]43 , and in our analyses these are resolved as sister to the resin-producing Detarioideae, although this relationship is weakly supported (Figs 2, 5). ...
Article
Full-text available
Detarioideae (81 genera, c. 760 species) is one of the six Leguminosae subfamilies recently reinstated by the Legume Phylogeny Working Group. This subfamily displays high morphological variability and is one of the early branching clades in the evolution of legumes. Using previously published and newly generated sequences from four loci (matK-trnK, rpL16, trnG-trnG2G and ITS), we develop a new densely sampled phylogeny to assess generic relationships and tribal delimitations within Detarioideae. The ITS phylogenetic trees are poorly resolved, but the plastid data recover several strongly supported clades, which also are supported in a concatenated plastid + ITS sequence analysis. We propose a new phylogeny-based tribal classification for Detarioideae that includes six tribes: re-circumscribed Detarieae and Amherstieae, and the four new tribes Afzelieae, Barnebydendreae, Saraceae and Schotieae. An identification key and descriptions for each of the tribes are also provided.
... The majority of monodominant forests in the Neotropics and Africa are composed of species in the legume family (Leguminosae), but compared with Africa and Asia, Neotropical monodominant-forming forest taxa are more diverse; this is true whether considering all families or the diversity of subfamilies and tribes among legume dominants (Table 1) [3,[5][6][7]. African monodominant tropical forests are nearly always characterized by legume species in the tribe Amherstieae of the subfamily Detarioideae, many of which occur in the monophyletic and African endemic Berlinia Clade (Table 1) [8,9]. A number of genera and species (even those not forming monodominant forest formations) within the Berlinia Clade have a tendency to form monotypic stands or are noted as often becoming gregarious where found [10]-thus, this African clade may have a propensity for forest ecological dominance. ...
Article
Full-text available
Many tropical wet forests are species-rich and have relatively even species frequency distributions. But, dominance by a single canopy species can also occur in tropical wet climates and can remain stable for centuries. These are uncommon globally, with the African wet tropics supporting more such communities than the Neotropics or Southeast Asia. Differences in regional evolutionary histories are implied by biogeography: most of Africa's mono-dominance-forming species are Amherstieae-tribe legumes; monodominance in Neotropical forests occur among diverse taxonomic groups, often legumes, but rarely Amherstieae, and monodominance in Southeast Asian forests occurs mostly among Dipter-ocarpaceae species. African monodominant forests have been characterized ecologically and taxonomically, but their deep-time history is unknown despite their significant presence and bottom-up ecological influence on diversity. Herein we describe fossil leaflets of Engler-odendron mulugetanum sp. nov., an extinct species of the extant genus Englerodendron (Berlinia Clade, Amherstieae, Detarioideae) from the 21.73 Ma Mush Valley site in Ethiopia. We also document a detailed study of associated legume pollen, which originate from a single taxon sharing characters with more than one extant descendant. Taxonomically, the pollen is most comparable to that from some extant Englerodendron species and supports a likely affiliation with the Englerodendron macrofossils. The Mush Valley site provides the first fossil evidence of a monodominant tropical forest in Africa as represented by leaflets and pollen. Previous studies documented >2400 leaves and leaflets from localities at six PLOS ONE PLOS ONE | https://doi.org/10.1371/journal.pone.
... All PCR amplification reactions were performed as described by Babineau, Gagnon & Bruneau (2013). Plastid and nrDNA regions were amplified according to previously published PCR cycling protocols as described by Estrella et al. (2014) for ITS and Wojciechowski et al. (2004) for the matK region. All PCR amplification products were submitted to Genome Quebec (Montreal, Canada), where they were purified and sequenced with Big Dye Terminator 3.1 chemistry on an ABI 3730xl DNA Analyzer (Applied Biosystems, Carlsbad, CA, USA). ...
Article
Full-text available
Annual Astragalus spp. are important elements in the floras of Mediterranean areas of Africa, Europe and the deserts of South-West Asia. By reconstructing phylogenetic trees including c. 75% of Old World annual Astragalus spp. and representatives of allied perennial taxa, we sought novel evolutionary insights regarding the phylogenetic position and origins of these species. Our analyses of nuclear ribosomal DNA internal transcribed spacer and plastid trnK/ matK sequences included 171 taxa of Astragalus plus nine outgroup taxa to recover the main clades in the genus. Character mapping for five selected morphological characters was conducted on the trees obtained, with a focus on the annual vs. perennial habit. Annual Astragalus spp. have originated in eight clades independently. Although the annual habit is the derived state in most clades recovered, at least two reversions to perenniality are found: one among the early branching lineages and another in the New World Astragalus clade. Our results suggest that the annual strategy has been an important force of selection leading to speciation in several clades of Astragalus and in different geographical areas. Geological events that have favoured the establishment of the annual habit in xeric habitats have probably also favoured the radiation of cushion-forming plants in the adjacent mountainous regions.
... Foug ere-Danezan et al., 2007;Bruneau et al., 2008;Estrella et al., 2014; M. de la Estrella et al., unpublished) ...
Conference Paper
Tropical rain forests are one of the most biologically diverse habitats on the planet. Yet their description and the understanding of their evolutionary history are far from complete. West Central Africa represents the area of greatest biodiversity richness in tropical Africa and has the highest percentage of untouched pristine forest in all of Africa and Madagascar. Our study is focused on the recently established subfamily Detarioideae (Leguminosae), which are the dominant tree components of West Central African forests. Detarioideae includes 81 genera (and approximately 750 species), 58% of which occur in Africa and Madagascar (with 329 species). We attempted to generate the first fully sampled phylogeny, including genera never sequenced before, using a phylogenomics approach (sequence capture). Our bait design includes 283 nuclear genes (359, 222 bp) and we aimed to obtain the plastid genome from the off target reads. Here we compare our results with the previous reconstructed the biogeographical history of the group and we have investigated biome shifts within the subfamily. Implications of the origin of the group and its role in shaping African diversity are discussed.
... Foug ere-Danezan et al., 2007;Bruneau et al., 2008;Estrella et al., 2014; M. de la Estrella et al., unpublished) ...
Article
Full-text available
African tropical forests are generally considered less diverse than their Neotropical and Asian counterparts. By contrast, the Detarioideae is much more diverse in Africa than in South America and Asia. To better understand the evolution of this contrasting diversity pattern, we investigated the biogeographical and ecological origin of this subfamily, testing whether they originated in dry biomes surrounding the Tethys Seaway as currently hypothesized for many groups of Leguminosae. We constructed the largest time-calibrated phylogeny for the subfamily to date, reconstructed ancestral states for geography and biome/habitat, estimated diversification and extinction rates, and evaluated biome/habitat and geographic shifts in Detarioideae. The ancestral habitat of Detarioideae is postulated to be a primary forest (terra firme) originated in Africa–South America, in the early Palaeocene, after which several biome/habitat and geographic shifts occurred. The origin of Detarioideae is older than previous estimates, which postulated a dry (succulent) biome origin according to the Tethys Seaway hypothesis, and instead we reveal a post Gondwana and terra firme origin for this early branching clade of legumes. Detarioideae include some of the most dominant trees in evergreen forests and have likely played a pivotal role in shaping continental African forest diversity.
Article
Full-text available
On the basis of a new phylogeny of the Detarioideae, with a particular focus on Englerodendron Harms, Anthonotha P.Beauv. and related genera, the possible options for delimiting monophyletic genera are discussed. As a result, Isomacrolobium Aubrév. & Pellegr. and Pseudomacrolobium Hauman are synonymised under Englerodendron. The following 12 new combinations are formed within the expanded Englerodendron: E. brachyrhachis (Breteler). A key to identification of the 17 species now recognised within Englerodendron is presented.
Article
Premise of research. The Neotropical endemic Macrolobium is one of the most species-rich genera (ca. 75 species) within subfamily Detarioideae (Fabaceae, alternatively Leguminosae). Two sections distinguished by floral morphology have been recognized in the past. Although morphologically diverse, species within the genus share several characters, including a single well-developed petal in adaxial position. However, previous analyses based on plastid markers have suggested that the genus is not monophyletic. We produce the most densely sampled molecular phylogeny of Macrolobium and test the monophyly of the genus and the two sections. Methodology. We analyzed nucleotide sequence data from the nuclear ribosomal internal transcribed spacer (ITS) and plastid (matK, trnG) genomes using Bayesian and maximum likelihood analyses. Pivotal results. The combined analysis retains Macrolobium as a monophyletic genus, with two well-supported subclades corresponding to the two recognized sections. Macrolobium pendulum is the only species placed in a section different from its taxonomic treatment. The relationships recovered with the plastid markers differ slightly from the combined and ITS analyses, but without significant support. Conclusions. Macrolobium is shown to be a monophyletic genus and to contain two well-supported and morphologically defined sections with differing Amazonian and Andean/Central American distributions corresponding to the Gentry pattern. Species are also found to group partly according to habitat preferences and leaf morphology. Both sections contain groups of multijugate and unijugate species, and there appear to have been multiple shifts of this character.
Article
The Gilbertiodendron ogoouense species complex consists of 14 tree species. Eight species are here newly described and one is here reinstated: G. bambolense Burgt; G. breteleri Burgt; G. ebo Burgt & Mackinder; G. ecoukense (Pellegr.) Burgt; G. maximum Burgt & Wieringa; G. minkebense Burgt & Estrella; G. quinquejugum Burgt; G. scutatum Wieringa & Estrella and G. sulfureum Burgt. Five species in the complex were already recognised as accepted: G. brachystegioides (Harms) J. Léonard; G. klainei (Pierre ex Pellegr.) J. Léonard; G. newberyi Burgt; G. ngouniense (Pellegr.) J. Léonard and G. ogoouense (Pellegr.) J. Léonard. All 14 species are medium-sized or large trees of evergreen rain forest on well-drained or periodically inundated soil, at 0 – 900 m altitude. Each of the 14 species is distributed in one or more of the following Central African countries: Cameroon, Equatorial Guinea, Gabon, Congo (Brazzaville), Angola (Cabinda) and Congo (Kinshasa). Included are a key, a table comparing the main morphological characters, illustrations drawn to the same scale and distribution maps of all species. The conservation status of all species is assessed according to IUCN categories and criteria: G. maximum is classified as Critically Endangered, G. ebo and G. newberyi as Endangered, G. breteleri, G. klainei, G. minkebense, G. ngouniense and G. scutatum as Vulnerable, and the remaining six species as Least Concern.
Research
The Gilbertiodendron ogoouense species complex consists of 14 tree species. Eight species are here newly described and one is here reinstated: G. bambolense Burgt; G. breteleri Burgt; G. ebo Burgt & Mackinder; G. ecoukense (Pellegr.) Burgt; G. maximum Burgt & Wieringa; G. minkebense Burgt & Estrella; G. quinquejugum Burgt; G. scutatum Wieringa & Estrella and G. sulfureum Burgt.
Book
Full-text available
The first edition of Tropical Rain Forests: an Ecological and Biogeographical Comparison exploded the myth of 'the rain forest' as a single, uniform entity. In reality, the major tropical rain forest regions, in tropical America, Africa, Southeast Asia, Madagascar, and New Guinea, have as many differences as similarities, as a result of their isolation from each other during the evolution of their floras and faunas. This new edition reinforces this message with new examples from recent and on-going research. After an introduction to the environments and geological histories of the major rain forest regions, subsequent chapters focus on plants, primates, carnivores and plant-eaters, birds, fruit bats and gliding animals, and insects, with an emphasis on the ecological and biogeographical differences between regions. This is followed by a new chapter on the unique tropical rain forests of oceanic islands. The final chapter, which has been completely rewritten, deals with the impacts of people on tropical rain forests and discusses possible conservation strategies that take into account the differences highlighted in the previous chapters. This exciting and very readable book, illustrated throughout with color photographs, will be invaluable reading for undergraduate students in a wide range of courses as well as an authoritative reference for graduate and professional ecologists, conservationists, and interested amateurs.
Book
— We studied sequence variation in 16S rDNA in 204 individuals from 37 populations of the land snail Candidula unifasciata (Poiret 1801) across the core species range in France, Switzerland, and Germany. Phylogeographic, nested clade, and coalescence analyses were used to elucidate the species evolutionary history. The study revealed the presence of two major evolutionary lineages that evolved in separate refuges in southeast France as result of previous fragmentation during the Pleistocene. Applying a recent extension of the nested clade analysis (Templeton 2001), we inferred that range expansions along river valleys in independent corridors to the north led eventually to a secondary contact zone of the major clades around the Geneva Basin. There is evidence supporting the idea that the formation of the secondary contact zone and the colonization of Germany might be postglacial events. The phylogeographic history inferred for C. unifasciata differs from general biogeographic patterns of postglacial colonization previously identified for other taxa, and it might represent a common model for species with restricted dispersal.
Article
1. The distribution of species-groups of Primates, Sciuromorpha, Artiodactyla and Hyracoidea in West African High Forest is given in detail. 2. Four types of distribution are recognized: a) ambisilvan, groups which occur both east of the Niger and west of the Volta. b) central, groups which occur only east of the Niger. c) occidental, groups which occur only west of the Volta. d) interfluvial, groups which occur between the Niger and the Volta, and which may have their origin from groups of type a), b) or c). 3. The central fauna is much richer than the occidental, which, though it contains a small number of endemic groups, is nevertheless thought to have been derived originally from the former. 4. The interfluvial fauna is remarkably poor. The zoogeographic data tend to corroborate geological evidence for the periodic deterioration of the climate between the Niger and the Volta during the Quaternary, which probably deprived the area of its High Forest several times, either partially or completely. 5. Climatic fluctuation has resulted in the Dahomey Gap acting as an important isolating mechanism in the past. Today, it is probably less important than the Volta and Niger rivers.