ArticlePDF Available

Sectional rearrangement of arborescent clades of Croton (Euphorbiaceae) in South America: Evolution of arillate seeds and a new species, Croton domatifer

Authors:
  • Forest Products Laboratory, Madison, WI, USA

Abstract and Figures

Most of the arborescent Croton species in the New World were treated by Webster as belonging either to C. sect. Cyclostigma Griseb. or C. sect. Luntia (Neck. ex Raf.) G.L. Webster. The circumscription of C. sect. Cyclostigma has been treated recently. In this paper we focus on C. sect. Luntia, which was subdivided by Webster into two subsections, C. subsect. Matourenses and subsect. Cuneati. Using chloroplast trnL‐F and nuclear ITS DNA sequence data, morphology and a broader sampling of additional Croton lineages, we found that the two subsections are not closely related and form distinct monophyletic groups. Substantial morphological differences support the molecular data. A taxonomic recircumscription of the two subsections, elevated to sectional level, is proposed. A key and taxonomic revision, with new synonyms, is provided for C. sect. Cuneati; together with the description of a new species from the coastal mountains of Venezuela, Croton domatifer. The new species is the only one in the genus known to possess leaf domatia. We infer that species in the Cuneati clade have lost the typical Croton caruncle, and some of them have evolved a different, specialized type of aril. We hypothesize that the arillate seeds are an adaptation to dispersal by fish in the Orinoco and Amazon river basins.
Content may be subject to copyright.
1147
Riina & al. • Arborescent Croton from South AmericaTAXON 59 (4) • August 2010: 1147–1160
INTRODUCTION
Croton L. is the second largest genus in Euphorbiaceae,
with an estimated 1250 species worldwide (Govaerts & al.,
2000; Frodin, 2004; Berry & al., 2005). The majority of the
species are shrubs adapted to seasonally dry areas of the trop-
ics and subtropics, but there are a substantial number of tree
species on different continents that occur in generally moister
habitats in both lowland and montane regions. One of the main
reasons why Croton trees are ecologically important is because
of their role as pioneer species in disturbed forest vegetation in
the tropics. For example, on Barro Colorado Island (Panama),
Croton is among the most common pioneers, with 367 repro-
ductive-sized individuals recorded in a census of a 50-ha plot
(D alling & Brow n, 2009). This ecological char acterist ic makes
Croton trees ideal candidates for restoration of degraded for-
ests, such as the use of C. urucurana Baill. in Brazil (Lorenzi,
1992; Carrenho & al., 1997). Most of the arborescent Croton
species from the New World were treated by Webster (1993) as
belonging to one of two sections, either C. sect. Cyclostigma
Griseb. or C. sect. Luntia (Ne ck. ex Raf.) G.L. Webster. Webster
(1993, 2001) also recognized two subsections within C. sect.
Luntia and four within C. sect. Cyclostigma. Riina & al. (2009)
used nuclear ITS and chloroplast trnL-F DNA sequence data,
as well as morphology, to determine that Webster’s circum-
scription of C. sect. Cyclostigma was widely polyphyletic, with
species treated in that section emerging in at least nine separate
clades in a broader sampling of Croton species. Two of these
nine clades represented the two subsections of Croton sect.
Luntia sensu Webster (1993), namely C. subsect. Matourenses
G.L. Webster and subsect. Cuneati G.L. Webster. Both Berry
& al. (2005) and Riina & al. (2009) showed that these two
subsections or clades are not part of C. sect. Cyclostigma s.str.,
and that they are not sister to each other, but rather represent
two morphologically distinct and not closely related lineages
within a larger clade of New World species.
Webster (1993) defined C. sect. Luntia based on a set of
morphological characters (e.g., lepidote indumentum, leaves
biglandular, stipules entire, inflorescences terminal and of-
ten clustered, sepals valvate, stamens 10–15, styles multifid,
etc.), however he did not indicate any single synapomorphy
to support the section. The two subsections of C. sect. Lun-
tia, C. subsects. Cuneati and Matourenses, were separated
by Webster (1993) based on four morphological differences:
presence or absence of bisexual cymules, pistillate flowers
subsessile or long-pedicellate, chorisepalous or gamopetalous
pistillate calyx, and capsules oblong or oblate, respectively.
Although Webster indicated that C. sect. Luntia was entirely
Sectional rearrangement of arborescent clades of Croton
(Euphorbiaceae) in South America: Evolution of arillate seeds
and a new species, Croton domatifer
Ricarda Riina,1,2 Benjamin van Ee,1,3 Alex C. Wiedenhoeft,4 Alfonso Cardozo5 & Paul E. Berry1
1 University of Michigan Herbarium and Department of Ecology and Evolutionary Biology, 3600 Varsity Drive, Ann Arbor,
Michig an 48108 -2228 , U. S.A.
2 Real Jardín Botánico, CSIC, Plaza de Murillo 2, 28014 Madrid, Spain
3 Harvard University Herbaria, 22 Divinit y Ave, Cambridge, Massachusetts 02138, U.S.A.
4 Forest Products Laboratory, One Gifford Pinchot Drive, Madison, Wisconsin 53726 -2398, U.S.A.
5 Universidad Central de Venezuela, Facultad de Agronomía, Instituto de Botánica Argrícola, Apartado 4579, Maracay,
Edo. Aragua, Venezuela
Author for correspondence: Ricarda Riina, riina@umich.edu
Abstract
Most of the arborescent Croton species in the New World were treated by Webster as belonging either to C. sect.
Cyclostigma Griseb. or C. sect. Luntia (Neck. ex Raf.) G.L. Webster. The circumscription of C. sect. Cyclostigma has been
treated recently. In this paper we focus on C. sect. Luntia, which was subdivided by Webster into two subsections, C. sub-
sect. Matourenses and subsect. Cuneati. Using chloroplast trnL-F and nuclear ITS DNA sequence data, morphology and a
broader sampling of additional Croton lineages, we found that the two subsections are not closely related and form distinct
monophyletic groups. Substantial morphological differences support the molecular data. A taxonomic recircumscription of the
two subsections, elevated to sectional level, is proposed. A key and taxonomic revision, with new synonyms, is provided for
C. sect. Cuneati; together with the description of a new species from the coastal mountains of Venezuela, Croton domatifer.
The new species is the only one in the genus known to possess leaf domatia. We infer that species in the Cuneati clade have
lost the typical Croton caruncle, and some of them have evolved a different, specialized type of aril. We hypothesize that the
arillate seeds are an adaptation to dispersal by fish in the Orinoco and Amazon river basins.
Keywor ds
arillate seeds; Croton domatifer; Croton section Cuneati; Croton section Luntia; Euphorbiaceae; leaf domatia;
molecular phylogeny; wood anatomy
1148
TAXON 59 (4) • August 2010: 1147–1160Riina & al. • Arborescent Croton from South America
Neotropical, he mentioned that some African species, such as
C. mubango Müll. Arg. from western central Africa, could be
closely related.
Croton sect. Luntia subsect. Matourenses was originally
circumscribed to include only two species, the type, C. ma-
tourensis Aubl., and C. lanjouwensis Jabl. (Webster, 1993).
The latter name, however, has since been treated as a syn-
onym of the variable C. matourensis (Webster & al., 1999).
Berry & al. (2005) then showed that the previously unplaced
C. megalodendron Müll. A rg. is si ster to C. matourensis in their
analysis, and Riina & al. (2009) showed that three species for-
merly treated by Webster as members of C. sect. Cyclostigma
subsect. Palanostigma Mart. ex Baill. (C. chocoanus Croizat,
C. palanostigma Klotzsch, C. smithianus Croizat) are sister to
C. megalodendron (C. matourensis was not included in this
analysis).
Croton sect. Luntia subsect. Cuneati was originally circum-
scribed by Webster (1993) to include about ten Neotropical spe-
cies, the type being C. cuneatus Klotzsch, a name applied to a
widespread Amazonian riparian tree. Other lowland Amazonian
or Orinocoan species in the subsection included C. tessmannii
Mansf., C. monachinoensis Jabl., and C. subcoriaceus Jabl.
Webster also included some montane species in the subsection,
such as C. kaieteuri Jabl. and C. neblinae Jabl., both from the
Guayana Shield, C. pachypodus G.L. Webster (a mainly An-
dean species from Costa Rica to Bolivia), and C. poecilanthus
Urb. (endemic to the Luquillo mountains of Puerto Rico). Van
Ee & al. (2008), however, showed that C. poecilanthus belongs
to an early diverging clade of Croton that is not at all closely
related to C. cuneatus. Riina & al. (2009) similarly showed
that C. pachypodus belongs to another early diverging branch
within Croton, and hence is not a member of the same clade as
C. cuneatus either. Berry & al. (2005) were the first to suggest a
close relationship between C. cuneatus and C. malambo Karst.,
which was previously placed by Webster (1993) in the mainly
Old World section, C. sect. Tiglium (Klotzsch) Baill., due to its
nearly glabrous foliage. Most recently, Riina & al. (2009) placed
C. yavitensis Croizat in the same clade with C. cuneatus based
on molecular phylogenetic analyses. Finally, we are now treat-
ing C. kaieteuri, C. monachinoensis, and C. subcoriaceus as
synonyms of C. cuneatus as a result of our taxonomic revision
of the pertinent type specimens.
This study focuses on the phylogeny and taxonomy of
Croton subsect. Matourenses and subsect. Cuneati. These
groups are important because they include tree species that
are prominent in primary and secondary forests of the Neo-
tropics, and they may provide insights into the evolution of
reproductive and vegetative characters within Croton, such as
the loss or modifications of the seed caruncle, reduplication of
the pistillate calyx, and the types of leaf glands and trichomes.
Croton cuneatus and C. malambo are also important as me-
dicinal plants in South America (Webster & al., 1999; Salatino
& al., 2007; Suárez & al., 2003, 2004, 2005, 2006, 2008). A
taxonomic recircumscription of both subsections is proposed,
with emphasis on C. subsect. (now sect.) Cuneati, for which we
include an identification key and a taxonomic synopsis. Finally,
we describe and illustrate a new species of C. sect. Cuneati,
Croton domatifer, the only one in the genus known to possess
well-developed leaf domatia.
MATERIALS AND METHODS
Molecular sampling. —
Species with characters consistent
with Croton sect. Luntia subsections Matourenses and Cuneati
sensu Webster (1993) were targeted for molecular sampling.
Th ese were combine d with a selectio n of t axa rep resent ing nine
of the eleven clades identified by Berry & al. (2005). The nuclear
ribosomal ITS (ITS1, 5.8s, ITS2) and plastid trnL-F (trnL exon,
intron, and 3 inte rgenic spa cer) lo ci we re sequenced e mploying
the same methods as described in Berry & al. (2005). These two
loci have been used in all species-level phylogenies of Croton
to date (Berry & al., 2005; Van Ee & al., 2008; Cordeiro &
al., 2008; Riina & al., 2009; Van Ee & Berry, 2009). Two out-
group taxa, Astraea lobata (L.) Klotzsch and Brasiliocroton
mamoninha P.E. Berry & Cordeiro, were included following
the results of Berry & al. (2005). Our finalized taxon sampling
contains 41 taxa for which ITS and trnL-F sequence data are
available for all accessions except C. smithianus, for which only
ITS is available. Indels were noted but not scored for inclusion
in the phylogenetic analyses. The data matrix is archived in
Tre eBASE (St udy accession nu mber = S2577; Matrix a ccession
number = M4924), and sequences are deposited in GenBank
(Appendix). Two sequences were newly generated for this study,
ITS and trnL-F sequences of the new species C. domatifer.
Phylogenetic analyses. —
DNA sequences were edited
and assembled using the Staden Package v.2003.0b1 (Staden,
1996), and the n alig ned manu ally u sing M acCla de v.4.08 (M ad-
dison & Maddison, 2001). Incongruence between the nuclear
and chloroplast loci was evaluated with the incongruence
length difference (ILD) test (Farris & al., 1994) as implemented
in PAUP* v.4.0b10 (Swofford, 2002). The ILD test was con-
ducted using 1000 partition homogeneity replicates of 10 ran-
dom addition sequence replicates (RASR) each, tree bisection
and reconnection (TBR), holding one tree at each step, nchuck
= 100, and excluding uninformative and gapped characters.
Croton smithianus, for which only ITS sequence data is avail-
able, was excluded from the ILD test.
Maximum parsimony (MP) analyses were conducted in
PAUP*. We performed MP heuristic searches with 100 ran-
dom taxon addition replicates using TBR branch swapping,
multrees in effect, and not limiting the number of trees saved.
All characters were equally weighted, and gaps were treated
as missing data.
The g1 statistic (Hillis & Huelsenbeck, 1992) and the parsi-
mony permutation tail probability test (PTP) (Faith & Cranston,
1991) were conducted on the combined ITS and trnL-F dataset
for assessing the phylogenetic signal of the combined data. The
Templeton test (Templeton, 1983), as implemented in PAUP*
under a maximum parsimony framework and with same settings
as in t he MP an alyse s above, was use d to evalu ate Web ster’s sug-
gestion that Croton sect. Cuneati and sect. Luntia form a clade.
The best fitting likelihood models for each of the loci, as
well as for the combined data, were selected with Modeltest
1149
Riina & al. • Arborescent Croton from South AmericaTA XON 59 (4) • August 2010: 1147–1160
v.3.07 (Posada & Crandall, 1998) using the Aikaike Informa-
tion Criterion (AIC). Ten independent likelihood analyses of the
loci in combination and separately were conducted in GARLI
v.0.96b8 (Zwickl, 2006). Each likelihood analysis used the
program’s default settings, the substitution model selected by
Model test, and allowing GARLI to estimate the model parame-
ters. In the combined likelihood analyses, the data were treated
as a single partition. Maximum likelihood (ML) bootstrap (BS)
values were calculated from the combined and separate loci
from 100 likelihood replicates run in GARLI, also using the
default stop criterion settings. Each BS replicate consisted of
four search replicates. Maximum parsimony BS values of the
combined and separate loci were calculated in PAUP* from
1000 replicates of 100 RASR each, TBR, and nchuck = 100.
Bayesian posterior probabilities (PP) of the combined loci
were calculated in MrBayes v.3.1.2 (Huelsenbeck & Ronquist,
2001) from two Markov chain Monte Carlo (MCMC) analyses,
each consisting of four linked chains (heat = 0.2), 1,000,000
generations, and sampling every 50 generations. The burn-in
period was estimated by visual examination of the likelihood-
by-generation plot. After removing the trees from the burn-in
period PP values were obtained by computing a majority rule
consensus of the trees from both MCMC chains.
Wood anatomy. —
Wood samples from Croton domati-
fer and other species in C. sect. Cuneati were prepared for
microscopic observation and measured following standard
microtechnique protocols as in Berry & Wiedenhoeft (2004).
RESULTS
Datasets. —
The aligned length of the ITS dataset is 704
positions, of which 326 (46.3%) are variable and 237 (33.7%)
are parsimony-informative. The aligned trnL-F dataset is
1251 positions long, of which 279 (22.3%) are variable and
131 (10.5%) are parsimony-informative. A visual inspection
of the data matrices did not reveal any obviously ambiguously
aligned regions, therefore no portions were excluded. The ILD
test failed t o reject (P = 0.314) the hy pot hesis of no mean ing ful
conflict between the two loci given this sampling. The phylo-
genetic signal in the combined data matrix was significant (g1
= –0.73, P < 0.01, and P = 0.01 in the PTP test). In contrast, the
Templeton test was not significant (P = 0.1537–0.2350), and the
hypothesis that Croton sect. Luntia and sect. Cuneati form a
clade, given these data, cannot be rejected.
The MP search of the combined ITS and trnL-F dataset
resulted in a single most parsimonious tree (not shown, L =
1388, consistency index = 0.595 and retention index = 0.654),
with both indices calculated including parsimony uninforma-
tive characters. Modeltest, using AIC, selected SYM + I + G for
the ITS data, TIM + I + G for the trnL-F d ata, and GTR + I + G
for the combined data. The ML model selected for the ITS data
(SYM + I + G) specifies equal codon frequencies (statefrequen-
cies = equal), which is suggested to be a poor option (Zwickl,
2006). Therefore, the ITS ML search was also conducted allow-
ing GARLI to estimate the codon frequencies (statefrequencies
= estimate), and the results were compared.
Although the ITS and trnL-F loci were analyzed sepa-
rately as well as together using parsimony, likelihood, and
Bayesian methods, only the result of the Bayesian analysis
of the combined data is presented here (Fig. 1), and where
relevant the results of the separate analyses are discussed.
All ten independent likelihood analyses of the combined loci
recovered the same topology (not shown), suggesting that the
program has converged on the best, or nearly best, topology.
There were no differences in the topology of the Bayesian
analysis and the likelihood analyses, and no strongly supported
differences between these and the results of the parsimony
analysis. Likewise, the separate ITS and trnL-F analyses do
not differ from each other or the combined analysis in any
strongly-supported way, with the exception that the trnL-F
analysis recovered Croton sect. Cyclostigma apart from
C. sect. Cascarilla Griseb., a result discussed in greater detail
in Riina & al. (2009).
Phylogenetic patterns. —
Croton cuneatus and four other
species, C. yavitensis, C. domatifer, C. roraimensis Croizat, and
C. malambo, form a strongly supported monophyletic group
(100 % PP/98% ML/97% MP BS; f rom here on the Baye sian PP,
ML bootstrap, and the MP bootstrap values are given in that
order), now labeled sect. Cuneati in Fig. 1 (previously subsect.
Cuneati). Croton sect. Cuneati is separate and not sister to the
well-supported (100%/100%/100%) C. sect. Luntia clade (pre-
viously C. subsect. Matourenses), which contains C. skutchii
Standl., C. megalodendron, C. matourensis, C. palanostigma,
C. smithianus, and C. chocoanus (Fig. 1). Although the overall
Croton sampling in thi s work is reduce d and s kewed compa red
to previous publications, the Bayesian consensus tree (Fig. 1)
shows the same overall topology as in previous Croton phy-
logenies in which two reciprocally monophyletic subgenera
are recovered. Also, New World clade 2 goes sister to the Old
World clade, and together these go sister to New World clade 1
(sensu Berry & al., 2005). Both C. sects. Cuneati and Luntia
belong to New World clade 2, equivalent to clades C-4 through
C-11 of Berry & al. (2005).
Croton sect. Cuneati is recovered as sister to C. sect. Cleo-
dora (Klotzsch) Baill. with weak support (77%/74%/<50%),
and C. sect. Luntia is recovered with moderate support
(100%/85%/58%) sister to a clade including C. sect. Lampro-
croton (Müll. Arg.) Pax and sect. Geiseleria (Klotzsch) Baill.
(Fig. 1). The C. sect. Luntia clade is formed by two sister sub-
clades, the first (top) with the entire and penninerved-leaved
C. matourensis sister to the lobed and palminerved-leaved
group of C. palanostigma, C. smithianus, and C. chocoanus.
The second, (bottom) subclade consists of C. megalodendron
and C. skutchii, both of which are entire and penninerved-
leaved species and are morphologically more similar to C. ma-
tourensis (Fig. 1). The species relationships within C. sect.
Cuneati are not as well resolved as in C. sect. Luntia, and only
C. cuneatus and C. yavitensis togethe r for m a h ighly sup por ted
clade (100%/100%/100%). Within C. sect.Cuneati, the lone dry
forest species, C. malambo, and the montane forest species,
C. roraimensis, form a polytomy with a clade containing the
other three species sampled, namely C. domatifer, C. yaviten-
sis, and C. cuneatus (Fig. 1).
1150
TAXON 59 (4) • August 2010: 1147–1160Riina & al. • Arborescent Croton from South America
Fig. .
Bayesian consensus tree of combined trnL-F and ITS data. The numbers represent support values in the following order: Bayesian pos-
terior probability/Maximum likelihood bootstrap support/Maximum parsimony bootstrap support. Names on the right of vertical bars represent
the sectional or informal clade assignment of the species.
1151
Riina & al. • Arborescent Croton from South AmericaTA XON 59 (4) • August 2010: 1147–1160
DISCUSSION
Two distinct and separate clades represent the two subsec-
tions of Croton sect. Luntia sensu Webster (1993), and given
their disparate positions in the phylogram (Fig. 1) and their
morphological differences, we recognize them as distinct sec-
tions, C. sect. Cuneati (G.L. Webster) Riina & P.E. Berry and
C. sect. Luntia (Neck. ex Raf.) G.L. Webster (see Taxonomic
Treatment below). The results presented in Fig. 1 give strong
Bayesian PP support, but not strong ML or MP bootstrap sup-
port, for C. sects. Cuneati and Luntia not being sister to each
other. However, the results of Berry & al. (2005) do provide
that support, further indicating that these clades are not closely
related to each other, unlike Webster’s (1993) hypothesis that
they were part of the same section. This pattern is also sup-
ported by clear morphological differences between the two
sections, which we describe in more detail below.
Croton sect. Cuneati is supported by two morphological
synapomorphies, namely loosely crenate to dentate leaf mar-
gins with sessile to shortly stipitate discoid glands in the si-
nuses, and the loss of the typical Croton caruncle (Fig. 2A–E).
All of these species effectively lack a caruncle on their seeds,
which otherwise is characteristic of almost all Croton species,
including all members of C. sect. Luntia. Croton roraimensis
and C. yavitensis both have unusually large seeds for the genus
(1.5–2.5 cm diam.), and C. domatifer has stellate-fasciculate
trichomes with unusually long radii and is the only species
known in the genus that has evident domatia-like structures on
the undersides of the leaves. Lastly, C. cuneatus and C. yavi-
tensis are strongly supported (100%/100%/100%) as sister
species , as previously reported by Riina & al. (2009). These are
both lowland riparian species, with seeds that have an arillate
structure that may be adapted to attract fish (Murillo, 1999)
(Fig. 2A–C). If this observation is confirmed by future studies,
it will indicate an interesting evolutionary shift of secondary
seed dispersal in Croton from ant to fish dispersal (nearly all
species of Croton have explosively dehiscent capsules that
initially scatter the seeds close to the mother plant). Additional
taxon sampling and a better resolved phylogeny is still needed
to identify the evolutionar y trends in seed morphology in this
group and the evolution of an elaborate aril that could be an
adaptation to riparian habitats and seasonally inundated low-
land forests (Fig. 2A–C). Other genera of Euphorbiaceae, such
as Hevea and Mabea, have b een documented to have seeds that
are eaten by fish in the same habitats, although these generally
lack an arillate structure (Kubitzki & Ziburski, 1994).
Based on its similar habitat and morphology to C. cu-
neatus, we hypothesize that C. tessmannii also belongs in
C. sect. Cuneati. Moreover, we predict that it will cluster with
C. cuneatus and C. yavitensis once molecular data is obtained
given that it is very similar morphologically to C. cuneatus,
the main difference being the seed size and the aril size and
shape of specimens of C. tessmannii. Two species placed by
Webster (1993) in his C. sect. Luntia subsect. Cuneati (here
re-ranked as C. sect. Cuneati), namely C. pachypodus and
C. poecilanthus, are confirmed here as belonging to C. subg.
Moacroton, a finding that is consistent with the results of Van
Ee & al. (2008) and Riina & al. (2009). Croton sect. Cuneati
shares several morphological features with C. sections Geisel-
eria and Corylocroton G.L. Webster. Although not closely
related (Fig. 1), these three sections share bifid styles, stellate
to lepidote trichomes, a pair of glands at the base of the leaf
blade, and the presence of glands in the sinuses of the leaf mar-
gins. All these shared character states are symplesiomorphic,
therefore they do not indicate close phylogenetic relationship.
Many species of C. sect. Geiseleria are annual herbs, but the
shrubby species can be distinguished from the other two sec-
tions by having unequal sepals in the pistillate flowers. Only
two species of C. sect. Corylocroton, C. caracasanus Pittier
and C. beetlei Croizat, occur in South America, and these can
be distinguished from those of C. sect. Cuneati by their carun-
culate seeds, very reduced pistillate sepals that are never re-
flexed in fruit, and epipetiolar glands at the base of the lamina
in the adaxial position. Our phylogenetic results indicate that
the presence of glands in the sinuses of the leaf margins is a
character that arose independently at least four times in Cro-
ton, three times in the Neotropical sections, C. sects. Cuneati,
Corylocroton (see Van Ee & al., 2008), and Geiseleria, and
one or more times on Madagascar. All Old World species of
Croton sampled to date form a monophyletic clade (Berry &
al., 2005; Van Ee & al., 2008; Riina & al., 2009), and within
Fig. .
Seeds and aril morphology of several members of Croton
sect. Cuneati and other Croton species.
A,
C. cuneatus, Va squ e z 7472
(AMAZ), showing ventral (bottom two) and dorsal (upper two) sides;
B,
C. tessmannii, Rimachi 11575 (AMAZ), showing dorsal (lower left)
and ventral (other three) sides;
C,
C. yavitensis, Beck 10126 (LPB),
showing ventral and dorsal sides;
D,
ecarunculate seeds of C. malam-
bo, Karsten s.n. (W), showing ventral sides;
E,
vestigial caruncle
of C. domatifer, Morillo 2601 (VEN), showing ventral sides;
F–G,
ventral side of seeds of C. ruizianus Mull.Arg., Ruíz & Pavón s.n.
(G-DC), and C. serratifolius Baill., Nee 4439 (USZ), showing more
typical Croton caruncles.
1152
TAXON 59 (4) • August 2010: 1147–1160Riina & al. • Arborescent Croton from South America
those all the Malagasy species form a monophyletic group
of their own (Berry & al., 2009). This finding contradicts
Webster’s hypothesis that C. sect. Luntia may have some Old
World members, such as the African C. mubango.
The placement of C. megalodendron and C. skutchii in
C. sect. Luntia (Fig. 1) confirms their close relationship to
C. matourensis, which was first discovered by Riina & al.
(2009). Based on morphological similarities, described in de-
tail in the following paragraph, we hypothesize that C. costatus
Kunth from Colombia and Andean Venezuela will be placed
with C. megalodendron and C. skutchii once it is sampled mo -
lecularly. The additional three species recovered in C. sect. Lun-
tia, namely C. chocoanus, C. palanostigma, and C. smithianus,
belong to what Webster treated as C. sect. Cyclostigma subsect.
Palanostigma Mart. ex Baill., and because C. palanostigma is
the type of C. subsect. Palanostigma, we synonymize it with
C. sect. Luntia. We predict that several more species previously
placed in C. subsect. Palanostigma, but not all, are actually
members of C. sect. Luntia rather than of C. sect. Cyclostigma
(see Taxonomic Treatment below).
In addition to the molecular characters, C. sect. Luntia is
defined by several f loral characters, including a robust and
strongly reduplicate calyx in the pistillate flowers, highly di-
vided styles, a low number of stamens (usually 10), and con-
spicuous linear-lanceolate bracts (3–4) subtending each cy-
mule (the bracts are deciduous as the inflorescence matures).
Unlike floral characters, foliar characters are not as uniform
across C. sect. Luntia, and there are two different types of
leaf morphology within the clade. There are species with pin-
nate venation and unlobed leaves, such as C. matourensis,
C. megalodendron, C. skutchii, and C. costatus, which also
lack laminar glands, and species with lobed and palmatinerved
leaves, which bear cup-shaped or patelliform glands on the
lamina (Fig. 3A–C). Within C. sect. Luntia, our molecular phy-
logeny (Fig. 1) supports an evolutionary trend from unlobed,
penninerved, eglandular leaves to deeply lobed, palminerved
leaves with numerous patelliform glands on both surfaces of
the lamina.
Although the Templeton test failed to reject the hypothesis
that C. sections Cuneati and Luntia are not sister clades, the
consistency of our phylogenetic pattern (Fig. 1) with previous
phylogenetic work on Croton (Berry & al., 2005; Van Ee &
al., 2008; Riina & al., 2009; Van Ee & Berry, 2009), and the
morphological differences between the two sections suggest
that they are not sister to each other. These two clades are in
a region of the Croton phylogeny (New World clade 2, Fig. 1)
with low resolution and support, which could explain the results
of the Templeton test. Additional taxon and molecular marker
sampling are needed to improve resolution and support along
the backbone of New World clade 2, which should support or
reject our hypothesis that C. sections Cuneati and Luntia do
not form a clade.
Fig. .
Diagnostic leaf char-
acters of Croton sect. Luntia
(A–C)
and sect. Cun eati
(D–F)
:
A,
entire and penninerved
leaves of C. costatus, Riina 1829
(VEN);
B,
lobed and palmately
veined leaves of C. pala-
nostigma, Riina 1492 (MICH);
C,
cup-shaped or patelliform
laminar glands (arrows) of
C. palanostigma, Mo rillo 9555
(VEN);
D,
leaves of C. cunea-
tus with subentire or loosely
dentate margins, Riina 1493
(MICH);
E,
leaves of C. yaviten-
sis with conspicuously dentate
margins;
F,
glandular leaf mar-
gin of C. tessmannii; the arrows
indicate two types of marginal
glands, ovoid glands on the
teeth and discoid glands in the
sinuses, Vás que z 11509 (USM).
1153
Riina & al. • Arborescent Croton from South AmericaTA XON 59 (4) • August 2010: 1147–1160
TAXONOMIC TREATMENT
Croton sect. Luntia (Neck. ex Raf.) G.L. Webster in Taxon 42:
804. 1993 – Luntia Neck. ex Raf. – Type: Luntia sericea
Raf., nom. illeg. (C. sericeus Lam., nom. illeg. C. ma-
tourensis Aubl.).
= Croton (sect. Cyclostigma Griseb.) subsect. Palanostigma
Mart. ex Baill. in Étude Euphorb.: 358. 1858 – Type: Croton
palanostigma Klotzsch.
Monoecious trees; indumentum lepidote or stellate-lep-
idote; leaves alternate, pinnately or palmately veined, entire
or tri-lobed (Fig. 3A–B), usually with sessile cup-shaped or
patelliform glands on one or both sides of the lamina (Fig. 3C)
and a pair of acropetiolar glands at the base of the lamina in
an abaxial position, stipules entire, deciduous; inflorescences
terminal, often clustered, unisexual or bisexual, bisexual cy-
mules sometimes present, conspicuous bracts (3–4) subtending
each cymule (more noticeable in very young inflorescences),
stamens 10–15; pistillate flowers with sepals entire, partially
united at the base, eglandular, reduplicate-valvate, styles re-
petitively divided, with 6–12 or more branches each; capsules
oblate, trilocular, seeds carunculate.
Included species: C. caryophyllus Benth., C. cearensis
Baill., C. costatus Kunth, C. chocoanus Croizat, C. grazielae
Secco, C. huitotorum Croizat, C. javarisensis Secco, C. kil-
lipianus Croizat, C. matourensis Aubl., C. megalodendron
Müll. Arg., C. nuntians Croizat, C. palanostigma Klotzsch,
C. perimetralensis Secco, C. pullei Lanj., C. skutchii Standl.,
C. smithianus Croizat, C. tonantinensis Jabl., C. uribei Croizat.
Necker (1790) first described Luntia as a genus. How-
ever, generic names published in his work are not accepted
as validly published (McNeill & al., 2006: Article 32.9). The
validly published Luntia Neck. ex Raf. therefore had to wait
until Rafinesque’s (1838) validation of the name. In this pub-
lication, Rafinesque made the new name Luntia sericea Raf.
for Lamarck’s (1786) C. sericeus Lam., which is illegitimate
and superfluous given that in the protologue, Aublet’s (1775)
earlier name C. matourensis Aubl. is cited.
There are still a number of taxonomic and nomenclatural
issues that need to be clarified for several species in Croton
sect. Luntia, and this will be done at a later date. However, a
report of a new synonym for C. skutchii follows. With fur-
ther study, C. skutchii may also prove to be conspecific with
C. costatus Kunth, which is the earliest name available.
Croton skutchii Standl. in Publ. Field Mus. Nat. Hist., Bot. Ser.
22: 86. 1940 – Type: Costa Rica. San José: vicinity of El
General, 670 m, A.F. Skutch 4377 (holotype, F!; isotypes,
K!, MO!, US!).
= Croton stipuliformis J. Murillo in Caldasia 21: 159, fig. 6.
1999 – Type: Colombia. Boyacá: Pauna, vereda Manote
Alto, 1250 m, 3 Nov 1975, H. Rueda s.n. (holotype, COL!).
Croton sect. Cuneati (G.L. Webster) Riina & P.E. Berry, stat.
nov. Croton subsect. Cuneati G.L. Webster in Taxon 42:
804. 1993 – Type: C. cuneatus Klotzsch.
Monoecious trees; indumentum lepidote, stellate-lepidote,
and stellate; leaves alternate, pinnately or palmately veined,
margin loosely crenate or sinuous, crenate or dentate, usually
with sessile or shortly stipitate discoid glands in the sinuses,
and sometimes less obvious ovoid glands on the teeth (Fig.
3D–F), and a pair of acropetiolar glands at the base of the
lamina in an abaxial position, stipules entire, deciduous; in-
florescences terminal, often clustered, unisexual or bisexual,
bisexual cymules sometimes present, styles 2-fid to 4-fid; sta-
mens 10–20; pistillate f lowers with sepals entire and eglandu-
lar, free, valvate; capsules oblong, trilocular, rarely bilocular,
seeds ecarunculate (Fig. 2D) or with a vestigial caruncle (Fig.
2E), sometimes with a distinctive aril (Fig. 2A–C).
Key to the species of Croton sect. Cuneati
1. Leaf blades broadly ovate, palmately 3–5-veined at the
base, less than twice as long as broad, indumentum of
dense ferrugineous appressed-stellate to rosulate tri-
chomes, with tufts of longer trichomes present in the axils
of abaxial leaf veins and forming domatia; in cloud forests
of the Venezuelan Coastal Cordilleras . . . . .C. domatifer
1. Leaf blades elliptic or ovate, penninerved or sometimes
3-veined at the base, more than twice as long as broad,
indumentum dense to very sparse (subglabrous), of lepi-
dote or stellate (but not ferrugineous) trichomes, domatia
absent; in lowland Amazonian forests, dry coastal forests,
or montane forest of the Guayana Shield . . . . . . . . . . . . 2
2. Le af bl ades nearly glabr ous, i f tr ichomes p resent, th en ve ry
sparse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Leaf blades sparsely to densely pubescent, at least on the
lower surface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
3. Leaf blades narrowly elliptic to lanceolate, apex acumi-
nate, base cuneate, petioles less than 1/5 the length of
the blade; pistillate flowers subsessile, ovary stellate to-
mentose, fruits and seeds unknown; upland forests of the
Guayana Shield in SE Venezuela and SE Colombia . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C. icabarui
3. Leaf blades ova te, ap ex acute, base rounde d to subcordat e,
petioles 1/4–1/3 the length of the blade; pistillate flowers
with pedicels 5–10 mm long, sepals of pistillate flowers
reflexed in fruit, capsules glabrous, seeds without aril or
caruncle; dry Caribbean coastal forests of NE Colombia
and NW Venezuela . . . . . . . . . . . . . . . . . . . . C. malambo
4. Young growth and undersides of leaves with stellate to
stellate-lepidote trichomes (becoming somewhat glabres-
cent); seeds 13–15 × 9–10 mm, aril with dissected margin
covering half or most of the ventral side of the seed, and the
upper part of the dorsal side (Fig. 2C); seasonally flooded
lowland vegetation (forests and savannas) in the Amazon
and Orinoco basins . . . . . . . . . . . . . . . . . . . .C. yavitensis
4. Young growth and undersides of leaves with lepidote
trichomes; seeds 6–9 × 46 mm or, if larger (15–25 ×
10–15 mm), then lacking an aril; montane forests of the
Gu ayana Shield as well a s seasonal ly f looded riverine for-
ests of the Amazon and Orinoco basins . . . . . . . . . . . . . 5
5. Lowland tress (<500 m elevation) in seasonally flooded
riverine forests of the Amazon and Orinoco basins; leaf
1154
TAXON 59 (4) • August 2010: 1147–1160Riina & al. • Arborescent Croton from South America
blades 3–5 times as long as wide; seeds ecarunculate but
with a ventral aril . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
5. Trees of moist montane forests of the Guayana Shield
(>1000 m elevation); leaf blades 2–3 times as long as wide;
mature seeds (not known in C. neblinae) 15–25 mm long
× 10–15 mm thick, seeds ecarunculate but with a broad
ventral scar lighter than the brown seed coat. . . . . . . . . 7
6. Leaf margin serrulate to subentire; aril with an entire
margin covering most of the ventral side of the seed (Fig.
2A); widespread in the Amazon and Orinoco basins (incl.
Guyana, French Guiana, and Suriname) . . . C. cuneatus
6. Leaf margin serrate; aril with an irregularly dissected
margin covering the upper part of the ventral side of the
seed (Fig. 2B); restricted to the Colombian and Peruvian
Amazon. . . . . . . . . . . . . . . . . . . . . . . . . . . .C. tessmannii
7. Leaf margins serrulate, petioles 2–4 cm long; slopes of
tepuis of eastern Venezuela and western Guyana . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. roraimensis
7. Leaf margins entire, petioles 6–10 cm long; only known
from the Sierra de la Neblina along the Venezuelan-Bra-
zilian border . . . . . . . . . . . . . . . . . . . . . . . . . .C. neblinae
Synopsis of the species of Croton sect. Cuneati
1. Croton cuneatus Klotzsch in London J. Bot. 2: 49. 1843 –
Type: Guyana. Roraima expedition, 1839, Richard Schom-
burgk s.n. (lectotype, designated here, K!; isolectotypes, B
destroyed, F-neg. 5079!, G!); Brazil. On the A mazon r iver,
1836, E.F. Poeppig 2593 (syntype, G!); Brazil. Rio Negro:
in sylvis Japurensibus, C.F.P. Martius s.n. (sy ntype, M
0089073!).
= Croton surinamensis Müll. Arg. in Linnaea 34: 82. 1865
Croton cuneatus Miq. in Linnaea 21: 477. 1848, non
C. cuneatus Klotzsch (1843) – Type: Suriname. Fr. W.R.
Hostmann 1094 (lectotype, designated here, G!; isolecto-
types, B destroyed, F-neg. 5184!, K!, G-DC!); Suriname.
A. Kappler 1505 (syntype, G!); Suriname. H.C. Hocke s.n.
(syntype, U?).
= Croton martii Müll. Arg. in Martius, Fl. Bras. 11(2): 88.
1873 C. martii var. latifolius Müll. Arg. in Martius, Fl.
Bras. 11(2): 88. 1873 – Type: Brazil. Pará: in silvis prope
Pará, C.F.P. Martius s.n. (lectotype, designated here, M
0089098!).
= Croton martii Müll. Arg. var. longifolius ll. Arg. in
Martius, Fl. Bras. 11(2): 88. 1873 – Type: Brazil. Pará:
prope Pará, C.F.P. Martius s.n., (lectotype, designated
here, M 0089099!).
= Croton mimeticus S. Moore in Trans. Linn. Soc. London
4: 453. 1895 – Type: Brazil. Matto Grosso: Villa Maria,
Dec 1891, S. Moore 843 (lectotype, designated here, BM;
isotype, B destroyed, F-neg. 5134!).
= Croton kaieteuri Jabl. in Mem. New York Bot. Gard. 12(3):
155. 1965 – Type: Guyana. Kaieteur Plateau, along Potaro
River above Kaiatuk, 10 May 1944, B. Maguire & D.B.
Fanshawe 23352 (holotype, NY!; isotypes, GH!, S!, US!,
VEN!).
= Croton monachinoensis Jabl. in Mem. New York Bot.
Gard. 12: 157. 1965 – Type: Venezuela. Bolívar: Río Sua-
pure, 17 Jan 1956, J. J. Wurdack & J. Monachino 41251
(holotype, NY!; isotypes, GH!, US!, VEN!).
= Croton subcoriaceus Jabl. in Mem. New York Bot. Gard.
12(3): 156. 1965 – Type: Venezuela. Amazonas: Río Ori-
noco, just below the mouth of Caño Yapacana, 125 m, 17
Mar 1953, J.J. Wurdack & L.S. Adderley 43028 (holotype,
NY!; isotypes, NY!, S!).
= Croton bilocularis J. Murillo in Caldasia 21: 156. 1999 –
Type: Colombia. Amazonas: Quebrada el Achote, af luente
del río Caquetá, frente a Peña Roja, Feb 1992, J. Murillo
& A. Matapí 85 (holotype, COL!; isotype, COAH!).
After examining the types of C. martii var. latifolius and
var. longifolius, and designating that of the former as lectotype
of the species name, we concluded that C. martii and both its
varieties should be synonymized with C. cuneatus. Croton
bilocularis was described as differing from C. cuneatus in hav-
ing bilocular capsules (Murillo, 1999). However, we found that
bilocular capsules are also present in some specimens of typical
C. cuneatus, and sometimes there are individuals with both
bilocular and trilocular capsules on the same inflorescences.
Murillo (1999) also indicated that C. bilocularis differed from
C. cuneatus in having the abaxial surface of the lamina covered
by a dense lepidote indumentum without spaces between the
scales. A dense lepidote indumentum is well within the range
of variation of this character in C. cuneatus, however, and we
could not find other significant differences between them.
Croton cuneatus is widespread in seasonally flooded ri-
parian forests of the Amazon and Guayana regions, including
Venezuela, Guyana, Suriname, French Guiana, Brazil, Colom-
bia, Ecuador, Peru, and Bolivia.
2. Croton icabarui Jabl. in Mem. New York Bot. Gard. 12(3):
158. 1965 – Type: Venezuela. Bolívar: región de los ríos
Icabaru, Hacha, y cordillera sin nombre a 280° de las cabe-
ceras del río Hacha, 450–850 m, 9 Jan 1956, A.L. Bernardi
2853 (holotype, NY!; isotype, VEN!).
This species is known from the type locality in southeast-
ern Venezuela and from one other locality in southeastern Co-
lombia (Schultes & Cabrera 19469, COL, U, from the Vaupés/
Caquetá border).
3. Croton malambo H. Karst . in Li nnae a 28: 418. 1856 Typ e:
Colombia. Bolívar: Cartagena, H. Karsten s.n. (lectotype,
designated here, LE!; isolectotype, W!).
This species is distributed in the Guajira region of north-
western Venezuela (Estado Zulia) and the northeastern coastal
region of Colombia (Departamentos Bolívar, Atlántico, Mag-
dalena, and La Guajira).
4. Croton neblinae Jabl. in Mem. New York Bot. Gard. 12(3):
155. 1965 – Type: Venezuela. Amazonas: Cerro de la
Neblina, Río Tatua: Cano Grande, SSE of Cumbre Camp,
26 Dec 1957, B. Maguire 42529 (holotype, NY!; isotype,
MO!).
Th is spec ies is only known f rom the ty pe locality in south-
ernmost Venezuela along the border with Brazil.
1155
Riina & al. • Arborescent Croton from South AmericaTA XON 59 (4) • August 2010: 1147–1160
5. Croton roraimensis Croizat in Bull. Torrey Bot. Club 67:
290. 1940 – Type: Venezuela. Bolívar: Mount Roraima,
southwestern slopes, about 7400 feet, 6 Jan 1939, A. Pinkus
122 (holotype, A!; isotypes, F!, NY!, US!).
= Croton roraimensis var. subinteger Steyerm. in Fieldiana,
Bot. 28: 315. 1952 – Type: Venezuela. Bolívar: Ptari-tepui,
along base of east-facing high sandstone bluff, 2410–2450
m, 7 Nov 1944, J.A. Steyermark 59925 (holotype, F!).
This species is an occasional tree of wet montane forests
of the Guayana region (Venezuela and Guyana).
6. Croton tessmannii Mansf. in Notizbl. Bot. Gart. Berlin-
Dahlem 9: 264. 1925 – Type: Peru. Loreto: Ucayali River,
f looded areas, Yarina Cocha, 150 m , 20 Nov 1923, G. Tess-
mann 3438 (holotype, B, destroyed; lectotype, designated
here, G!; isolectotypes, NY!, S!). Peru. Loreto: upper
Ucayali, Cumaria, s.d., G. Tessmann 3434 (paratypes: B
destroyed, F-neg. 5186!, F!, NY!).
In several major herbaria, the name Croton tessmannii
has mistakenly been applied to many specimens of C. pachy-
podus, which is a medium-elevation montane forest species
that is widespread along the foothills of the Andes and also
occurring in Panama and Costa Rica. Croton pachypodus can
be distinguished from C. tessmannii by its large fruits (the
largest fruit of all Neotropical Croton species, 3–5 cm diam.),
petiolar glands acropetiolar and in adaxial position (vs. glands
on the abaxial side in C. tessmannii), 2-fid styles (vs. 4-fid
in C. tessmannii), and axillary inf lorescences (vs. a terminal
cluster of inflorescences in C. tessmannii). Croton pachypodus
was erroneously described as having 4-fid styles (Webster &
Huft, 1988), but, like its close relatives C. megistocarpus J.A.
González & Poveda and C. jorgei J. Murillo, it has bif id styles.
Croton tessmannii is known quite locally from riparian
lowland forests in Loreto, Peru.
7. Croton yavitensis Croizat in J. Arnold Arbor. 26: 189.
1945 – Type: Venezuela. Amazonas: Yavita, L.W. Williams
14029 (holotype, A!; isotypes, G!, US!, VEN!).
This species occurs in seasonally f looded riparian forests
of the Amazon and Orinoco basins in Venezuela (Amazonas),
Colombia (Guainía), Brazil (Rondônia, Goiás, Tocantins, Mato
Grosso), and northern Bolivia (Beni, La Paz, Pando, Santa
Cruz). Individuals from the southernmost part of the species
distribution exhibit more stellate-lepidote trichomes (50% web-
bing of the trichome radii), whereas individuals around the type
locality (Amazonas, Venezuela) show more stellate trichomes
(<5% of webbing of the trichome radii).
8. Croton domatifer Riina & P.E. Berry, sp. nov. – Type:
Venez uela. A rag ua: Pa rque Naciona l Henri Pittier, 1700 m ,
selva nublada superior, 1 Jul 1992, A. Cardozo, D. Hidalgo
& R. Hidalgo 1992 (holotype, MY!; isotypes, MICH!,
MY!, VEN!). Figures 2E and 4.
Arbor 5–25-metralis Crotoni roraimensis affinis, sed
differt: foliis cordato-ovatis margine haud crenato-dentatis,
trichomatibus porrectis 0.7–1.0 mm longis, domatiis abaxil-
laribus praesentibus.
Monoecious tree 5–25 m high, indumentum of stellate,
rosulate, to stellate-lepidote trichomes; latex from the trunk
and young branches dar k orange to red dish. Stipules cupul ate to
cucullate, 1.8–2.5 × 1.8–2.5 mm, persistent on young branches.
Leaves alternate, petioles 6–13 cm long, lamina broadly ovate
to cordate, 9–19 × 5–16 cm, apex acute, base usually cordate
or le ss com monly rounded; marg in i rreg ularly crenat e-dentate,
minute patelliform glands usually present in the sinuses, vena-
tion palmate with 3–5 veins from the base and 4–7 lateral veins
per side of the midrib further up the lamina, the primary and
secondary veins raised on the abaxial surface; indumentum on
both side s of t he lamina of m atu re leaves sparse to d ens e, den ser
along the veins, trichomes appressed-stellate, stellate-lepidote
to rosulate, sometimes with a longer porrect central ray; vein
axils on the abaxial surface of the lamina and both sides of the
leaf base covered with dense clusters of rosulate, ferrugineous
trichomes with long rays 0.7–1.0 mm long, forming domatia;
petiola r gland s 2, acropetiolar, patellifor m, 0.6 –0.8 mm diam.,
sessile to shortly stipitate. Inflorescences terminal, 1–5 inf lo-
rescences per apical branch, erect, 10–20 cm long; young inflo-
rescences with basal bisexual cymules, mature inflorescences
bearing either pistillate or staminate flowers, rachis densely
covered by appressed-stellate trichomes. Staminate flowers
2–3 mm long, 3–5 mm wide; pedicels 3–11 mm long; sepals
5, ovate, 2.7–3.0 × 1.8–2.0 mm, slightly imbricate at the base,
abaxial surface glabrous and papillose, adaxial surface sparsely
villous, apex acute and densely villous at the tip, base truncate;
petals 5, oval-oblong, 2.0–2.8 × 1.0–1.6 mm, apex acute to
rounded, villous at the tip, glabrous and papillate on both sides,
margins sparsely villous; receptacle densely villous; stamens
10–11; filaments 0.5–0.6 mm long, densely villous; anthers
0.8–1.0 mm long. Pistillate flowers with pedicels 5–13 mm
long; sepals 5, valvate, broadly obovate, slightly imbricate in
young flowers, reflexed in fruit, 3.5–4.0 × 3.0–3.2 mm, apex
acute, adaxial surface pilose, abaxial surface covered with
sparse stellate trichomes; ovary globose, ± trilobed, 3.0–4.5 ×
2.5–5.0 mm, covered with a dense indumentum of light golden
stellate and stellate-lepidote trichomes; petals absent or reduced
to narrow g landular filaments; styles bif id. Fr uit s globose, 1.0 –
1.2 cm diam.; columella 6–7 mm long; seeds ovoid, 6.0–7.0 ×
4.0–5.0 mm, dorsally compressed, base ending in a pointed
tip, apex obtuse, aril reduced to a small, irregular appressed
structure above the hilum.
Wood anatomical description. —
Macroscopic features:
Growth rings not distinct with a hand lens. Wood white to light
brown, heartwood absent or not distinct. Basic specific grav-
ity: 0.38. Microscopic features: Growth increments indistinct
or absent. Wood diffuse-porous; vessels solitary and in radial
multiples of 2–4; 6 vessels/mm2 (Fig. 5A); helical thickenings
occasionally present in tails of vessel elements. Vessel outline
rounded; average tangential diameter 99 m. Vessel elements
737 μm long. Perforation plates exclusively simple. Intervessel
pits non-vestured, alternate; 11 μm diameter (Fig. 5F); with
rounded outline. Vessel-ray pits similar to intervessel pits in
size and shape, and vessel-ray pits larger than intervessel pits
and simple (Fig. 5E). Fibers non-septate; thin- to thick-walled;
without helical thickenings; angular outline; 1226 μm long.
1156
TAXON 59 (4) • August 2010: 1147–1160Riina & al. • Arborescent Croton from South America
Fig. .
Croton domatifer.
A,
flowering branch;
B,
petiolar glands;
C,
pistillate flower;
D,
staminate f lower;
E,
marginal leaf glands;
F,
stipule;
G,
stellate-lepidote and rosulate trichomes;
H,
rosulate trichome with a single, long porrect central ray;
I,
rosulate trichome with numerous long
rays, forming domatia on the abaxial surface of leaves in vein axils and at the base. Note: items E–I are not to scale.
1157
Riina & al. • Arborescent Croton from South AmericaTA XON 59 (4) • August 2010: 1147–1160
Fig. .
Collection localities of Croton domatifer in northern Venezu-
ela indicated by dots.
Fiber pits distinctly bordered (Fig. 5E). Fiber length to vessel
element length ratio 1.66. Paratracheal parenchyma scanty and
sparsely aliform. Apotracheal parenchyma diffuse to diffuse-
in-aggregate, with the diffuse-in-aggregate parenchyma in
confluent-like arrangement (Fig. 5B). Parenchyma strands
4-celled. Rays 1–2-seriate; 11 per linear mm; 228 μm in aver-
age height; tallest ten rays average 340 μm (Fig. 5D). Rays
composed of square, upright, and procumbent cells, upright
cells in marginal rows of more than four common (Fig. 5C).
Disjunctive ray cell end walls common, especially in upright
cells. Cr ystals rare to absent in bot h axial and ray pare nchy ma.
Wood never storied.
Distribution and habitat. —
Known from Aragua, Su-
cre, and Yaracuy states in northern Venezuela (Fig. 6). Croton
domatifer occurs in montane cloud forests of the Coastal Cor-
dillera, between 760 and 1850 m elevation.
Etymology. —
The specific epithet refers to the presence
of leaf
domatia
.
Additional specimens examined. —
VENEZUELA.
Aragua: Parque Nacional Henri Pittier, Pico Guacamaya
1850 m, 14 Apr 1990, A. Cardozo & E. Garcia 1372 (MY); 1
Fig. .
Light micrographs of the wood of Croton domatifer, Riina 1537 (VEN).
A, C–F,
transmitted light micrographs;
B,
stereomicrograph.
A,
transverse section showing relatively large vessels and thin walled fibers;
B,
transverse surface showing the diffuse-in-aggregate parenchyma
in confluent-like arrangement;
C,
radial section showing heterocellular rays;
D,
tangential section showing uniseriate rays;
E,
transverse section
showing ray-vessel pits ranging from the same as intervessel pits to larger and simple (unbordered), and distinctly bordered fiber pits indicated
by arrows;
F,
tangential section showing intervessel pits.
1158
TAXON 59 (4) • August 2010: 1147–1160Riina & al. • Arborescent Croton from South America
Jul 1990, A. Cardozo & al. 1490 (M Y); Parque Nacional Hen ri
Pittier, 1700 m, 23 Jun 1991, A. Cardozo & D. Hidalgo 1795
(MY); 1450–1500 m, 21 May 1992, A. Cardozo & al. 1983
(M ICH, M Y); Rancho G rande camino hacia Pico Gu acamayo,
1400 m, 4 Jan 1976, O. Huber 360 (DAV, VEN); entre Cumbre
de Rancho Grande y Pico Guacamayo, 1430 m, 28 Jan 1976,
O. Huber 378 (VEN); 1570 m, G. Markina 5 (MER); camino
hasta la Cumbre de Rancho Grande, 1340–1450 m, 23 Apr
2006, R. Riina & al. 1537 (VEN). Sucre: Península de Paria,
Cerro Humo, 62°36 W, 10 °41 N, 12 Jul 1972, K. Dumont &
al. VE-7578 (NY), G. Morillo 2601 (MY, VEN); entre Roma
y Santa Isabel, 1273 m, 2 Mar 1966, J.A. Steyermark 94900
(NY, VEN); 10°41–42 N, 62°36–37 W, 760–1000 m, 24 Feb
1980, J.A. Steyermark & al. 121742 (F, DAV, MO, VEN). Yar-
acuy: Sierra de Aroa, Cerro Negro, 1200–1800 m, 10°17 N,
69°01 W, 1–2 April 1980, R. Liesner & A. González 9905
(VEN); Río Cocorotico, arriba de San Felipe, 1150–1350 m,
14 Nov 1967, J.A. Steyermark & G. Wessels-Boer 100463
(DAV, NY, VEN).
Specimens from Sucre state in northeastern Venezuela
have smaller, ovate-lanceolate leaves with a rounded base in
comparison with the larger, broadly and deeply cordate leaves
of specimens from Aragua and Yaracuy states. The descrip-
tion of fruits and seeds is based on Morillo 2601 (VEN, MY),
Tab le .
Morphological and wood anatomical differences between Croton domatifer and several representatives of Croton sect. Cuneati.
C. cuneatus C. domatifer C. malambo C. roraimensis C. yavitensis
Stipules Linear-narrowly
triangular
Cupulate-cucullate Linear-lanceolate Linear-subulate
(asymmetric)
Linear-narrowly
triangular
Leaf base Cuneate Deeply cordate to
rounded
Slightly cordate to
rounded
Cuneate to rounded Cordate to rounded
Leaf margin Entire to crenate-
dentate
Crenate-dentate Loosely crenate Dentate Crenate to dentate
Leaf domatia Absent Present Absent Absent Absent
Stamens 16 10–11 18–20 12–15 10
Pedicel of pistillate flowers 1–2 mm 10–13 mm 17–20 mm 4–6 mm 1–2 mm
Ovary indument; trichomes Dense; lepidote Very dense; stellate to
stellate-lepidote
Glabrous (caducous) Dense; lepidote Very dense; stellate-
lepidote
Style 4-fid 2-fid 2-fid 2-fid 4-fid
Fruit length 13–16 mm 10–12 mm 10–11 mm 25–30 mm 20–25 mm
Caruncle Absent Absent Absent Absent Absent
Aril Present, conspicuous Present, incon-
spicuous
Absent Absent Present, conspicuous
Habitat Riparian lowland
forests
Montane forests Dry lowland forests Montane forests Riparian lowland
forests
Distribution Amazon and Orinoco
basins
Northern Venezuela,
Coastal Cordillera
Guajira region (Ven-
ezuela and Colombia)
Guayana Highlands Amazon basin
Vessels per square mm 7–13 6 21–43 9 9
Tangential vessel diameter 70–123 μm 99 μm 79–83 μm 114 μm 77 μm
Ray-vessel pit type Same as intervessel
pits
Same to smaller and
simple
Same as intervessel
pits
Same to unilaterally
coarse and compound
Smaller than interves-
sel pits
Intervessel pit diameter 10–11 μm11 μm6 μm 12 μm 12 μm
Fiber pitting Distinctly bordered Distinctly bordered Indistinctly bordered Indistinctly bordered Indistinctly bordered
Helical thickenings in
vessel elements
In tails In tails Absent Absent In tails
1159
Riina & al. • Arborescent Croton from South AmericaTA XON 59 (4) • August 2010: 1147–1160
the only collection with mature fruits. Croton domatifer is
remarkably distinctive from other species of Croton because
of the leaf domatia which consist of dense clumps of large
stellate-rosulate trichomes. Apparently, this is the first report
of leaf domatia in the genus.
Croton domatifer shows morphological affinities to the
species in Croton sect. Cuneati, and DNA sequence data
(trnL-F, ITS) support its position in this clade (Fig. 1). Table
1 shows the main morphological and wood anatomical differ-
ences and similarities between the new species and several
representatives of the Cuneati clade (C. cuneatus, C. yaviten-
sis, C. malambo, C. roraimensis). Species in C. sect. Cuneati
and C. domatifer share characters such as dentate to loosely
crenate leaf margins (sometimes subentire in C. cuneatus), dis-
coid or patelliform glands along the leaf margin, indumentum
of lepidote or stellate-lepidote trichomes, patelliform petiolar
glands, a very reduced caruncle or ecarunculate seeds, some
species have a non-carunculate aril covering a large area of
the ventral surface of the seed (Fig. 2A–D), and preferences
for mesic habitats such as riversides,
seasonally flooded
ev-
ergreen
forests
, and montane cloud forests. The only appar-
ent exception to this syndrome is C. malambo, which occurs
in dry forests. Croton domatifer displays a wood anatomical
pattern typical of mesic, arborescent Croton, consistent with
other members of the Cuneati clade (Wiedenhoeft, 2008). An
intriguing feature of the wood of C. domatifer is the presence
of diffuse-in-aggregate parenchyma in confluent-like bands,
which cannot be seen clearly with transmitted light micros-
copy, but is easily observed with a stereomicroscope or a hand
lens (Fig. 5B). Parenchyma of this type is not common, and
the IAWA list does not recognize it as a separate character
(I AWA Committee, 1989). The presence of distinctly bordered
fiber pits (Fig. 5E) in this new species is an important char-
acter linking it to C. sect. Cuneati, these distinctive pits have
been found in most species of C. sect. Cuneati examined to
date (C. cuneatus, C. domatifer, C. neblinae, C. roraimensis),
although they are not found in C. malambo and C. yaviten-
sis (Wiedenhoeft, 2008). These distinct pits are also found
in C. sampatik and C. poecilanthus (Wiedenhoeft, 2008), a
species once thought to be in C. sect. Cuneati (as subse ct.
Cuneati) by Webster (1993).
ACKNOWLEDGMENTS
We wish to thank Bruno Manara for the illustration of C. domati-
fer and Kandis Elliot for the distribution map. We are grateful to the
staff of A, AMAZ, BM, COAH, COL, DAV, F, G, G-DC, GH, K, LE,
LPB, M, MA, MER, MICH, MO, MY, NY, S, US, USM, USZ, VEN,
W, and WIS for giving us access to their collections and digital images,
to Carlos Reyes, Aurimar Magallanes, and Dumas Conde for their
help in the field, to Nicolas Fumeaux (G) for his invaluable help with
taxonomic and nomenclatural issues, and to two anonymous reviewers
for their comments, corrections, and useful suggestions. This study
was supported in part by the U.S. National Science Foundation (award
# DEB-0212481 to P. Berry) and by the International Association for
Plant Taxonomy (award to R. Riina).
LITERATURE CITED
Aublet, F. 1775 . Histoire des plantes de la Guiane Françoise, vol. 2.
London, Paris: Didot.
Berry, P.E., Hipp, A.L., Wurdack, K.J., Van Ee, B. & Riina, R.
2005. Molecular phylogenetics of the giant genus Croton and tribe
Crotoneae (Euphorbiaceae sensu stricto) using ITS and trnL-trnF
DNA sequence data. Amer. J. Bot. 92: 1520–1534.
Berry, P.E., Riina, R., Van Ee, B. & Haber, E. 2009. The radiation
of Croton (Euphorbiaceae) on Madagascar—how did this group
become so speciose and diversified there compared to other Old
World areas? Botany & Mycology 2009, Snowbird, Utah, July
25–29. Abstract, available at: www.2009.botanyconference.org.
Be rry, P.E . & W ieden hoeft, A.C . 2004. Micrandra inundata (Euphor-
biaceae), a new species with unusual wood anatomy from black-
water river banks in southern Venezuela. Syst. Bot. 29: 125–133.
Carrenho, R., Bononi, V.L.R. & Barbosa, L.M. 1997. Glomales em
áreas de recomposição de mata ciliar de Moji-Guaçu, SP, Brasil.
Hoehnea 24: 107–113.
Cordeiro, I., Berry, P.E., Caruzo, M.B. & Van Ee, B. 2008. Croton
laceratoglandulosus (Euphorbiaceae s.s.), a new glandular-stip-
ulate species from Brazil an Bolivia, and its systematic position
based on molecular analysis. Bot. J. Linn. Soc. 158: 493– 498.
Dalling, J.W. & Brown, T.A. 2009. Long-term persistence of pioneer
species in tropical rain forest soil seed banks. Amer. Naturalist
173: 531–535.
Faith, D.P. & Cranston, P.S. 1991. Could a cladogram this short have
arisen by chance alone? On permutation tests for cladistic struc-
ture. Cladistics 7: 1–28.
Farris, J.S., Kallersjø, M., Kluge, A.G. & Bult, C. 1994. Testing
significance on incongruence. Cladistics 10: 315–319.
Frodin, D.G. 2004. History and concepts of big plant genera. Ta x o n
53: 753–776.
Govaerts, R., Frodin, D.G. & Radcliffe-Smith, A. 2000. World c hec k-
list and bibliography of Euphorbiaceae (and Pandaceae), 4 vols.
Kew: Royal Botanic Gardens.
Hillis, D.M. & Huelsenbeck, J.P. 1992. Signal, noise, and reliability in
molecular phylogenetic analyses. J. Heredity 83: 189–195.
Huelsenbeck, J.P. & Ronquist, F. 2001. MrBayes: Bayesian inference
of phylogeny. Bioinformatics 17: 754 –755.
IAWA Committee. 1989. IAWA list of microscopic features for hard-
wood identification. I.A.W.A. Bull., n.s., 10: 219–332.
Kubitzki, K. & Ziburski, I. 1994. Seed dispersal in flood plain forests
of Amazonia. Biotropica 26: 30–43.
Lamarck, J. 178 6. Encyclopédie méthodique, Botanique, ed. 2. Paris:
Panckoucke.
Lorenzi, H. 1992. Arvores brasileiras: Manual de identificação e
cultivo de plantas arbóreas nativas do Brasil, vo l. 1. N ova O des sa:
Editora Plantarum.
Maddison, D.R . & Maddison, W.P. 2001. MacClade: Analysis of phy-
logeny and character evolution, version 4.03. Sunderland, Mas-
sachusetts: Sinauer.
McNeill, J., Barrie, F.R., Burdet, H.M., Demoulin, V., Hawksworth,
D.L., Marhold, K., Nicolson, D.H., Prado, J., Silva, P.C., Skog,
J.E., Wiersema, J.H. & Turland, N.J. (eds.). 2006. International
code of botanical nomenclature (Vienna Code): Adopted by the
Seventeenth International Botanical Congress Vienna, Austria,
July 2005. Regnum Vegetabile 146. Ruggell: Gantner.
Murillo, J. 1999. Composición y distribución del género Croton (Eu-
phorbiaceae) en Colombia, con cuatro especies nuevas. Caldasia
21: 141–166.
Necker, N.J.
179 0. Corollarium ad philosophiam botanicam Linnaei
spectans. Neoweda ad Rhenum [Neuwied]: apud Societatem Ty-
pographycam.
Posada, D. & Crandal l, K.A. 1998. MODELTEST: Testing the model
of DNA substitution. Bioinformatics 14: 817–818.
Rafinesque, C. S. 1838. Sylva telluriana. Philadelphia.
1160
TAXON 59 (4) • August 2010: 1147–1160Riina & al. • Arborescent Croton from South America
Riina, R., Berry, P.E. & Van Ee, B.W. 2009. Molecular phylogenetics
of the dragon’s blood Croton sect. Cyclostigma (Euphorbiaceae): A
polyphyletic assemblage unraveled. Syst. Bot. 34: 360–374.
Salatino, A., Salatino, M.L.F. & Negri, G. 2007. Traditional uses,
chemistry and pharmacology of Croton species (Euphorbiaceae).
J. Brazil. Chem. Soc.
18:
11
33
.
Staden, R. 1996. The Staden sequence analysis package. Molec. Bio-
technol. 5: 233–241.
Suárez, A.I., Blanco, Z., Compagnone, R.S., Salazar-Bookaman,
M.M., Zapata, V. & Alvarado, C.
2006.
Anti-inflammatory ac-
tivity of Croton cuneatus aqueous extract. J. Ethno-pharmacol.
105
:
99
101
.
Suárez, A.I., Blanco, Z., Delle Monache, F., Compagnone, R .S. &
Arvelo, F.
2004.
Three new glutarimide alkaloids from Croton
cuneatus. Nat. Prod. Res.
18:
421
426
.
Suárez, A.I., Compagnone, R.S., Salazar-Bookaman, M.M., Tillett,
S., Delle Monache, F., Di Giulio, C. & Bruges, G.
2003.
Antino-
ciceptive and anti-inflammatory effects of Croton malambo bark
aqueous extract. J. Ethno-pharmacol.
88
:
11
14
.
Suárez, A.I., Vásquez, L.J., Manzano, M.A. & Compagnone, R.S.
2005. Essential oil composition of Croton cuneatus and Croton
malambo growing in Venezuela. Flavour Fragr. J. 20: 611–614.
Suárez, A.I., Vásquez, L.J., Taddei, A., Arvelo, F. & Compagnone,
R.S. 2
008.
Antibacterial and cytotoxic activity of leaf essential
oil of Croton malambo. J. Essential Oil-Bearing Pl.
11
:
208
213
.
Swofford, D.L . 2002. PAUP*: Phylogenetic analysis using parsimony
(*and other methods), version 4.0b10. Sunderland, Massachusetts:
Sinauer.
Tem p l e ton, A . R . 1983. Phylogenetic inference from restriction endo-
nuclea se cl eavage sit e maps wit h parti cular refer enc e to t he evol u-
tion of humans and the apes. Evolution 37: 221–244.
Van Ee, B.W. & Berry, P.E. 2009. A phylogenetic and taxonomic re-
view of Croton (Euphorbiaceae s.s.) on Jamaica including the de-
scription of Croton jamaicensis, a new species of section Eluteria.
Syst. Bot. 34: 129–140.
Van Ee, B.W., Berry, P.E., Riina, R. & Gutiérrez Amaro, J.E. 2008.
Molecular phylogenetics and biogeography of the Caribbean-cen-
tered Croton subgenus Moacroton (Euphorbiaceae s.s.). Bot. Rev.
74:
132
165
.
Web st er, G .L. 1993. A provisional synopsis of the sections of the genus
Croton L. (Euphorbiaceae). Tax on 42: 793–823.
Web st er, G .L. 2001. Synopsis of Croton and Phyllanthus (Euphorbia-
ceae) in Western Tropical Mexico. Contr. Univ. Michigan Herb.
23: 353–388.
Webster, G.L., Berry, P.E ., Armbruster, W.S., Esser, H.-J., Gil-
lespie, L.J., Hayden, W.J., Levin, G.A., Secco, R.S. & Heald,
S.V. 1999. Euphorbiaceae. Pp. 74–228 in: Berr y, P.E., Yatsk ievych,
K. & Holst, B.K. (eds.), Flora of the Venezuelan Guayana, vol. 5.
St. Louis: Missouri Botanical Garden Press.
Webster, G.L. & Huft, M.J. 1988. Revised synopsis of Panamanian
Euphorbiaceae. Ann. Missouri Bot. Gard. 75: 1087–1144.
Wiedenhoeft, A.C. 2008. Tracking the phylogeny of the Crotoneae with
comparative wood anatomy.
Ph.D. Dissertation,
University of
Wisconsin-Madison, Madison.
Zwickl, D.J. 2006.
Genetic algorithm approaches for the phylogenetic
analysis of large biological sequence datasets under the maximum
likelihood criterion
. Ph.D. Dissertation, The University of Texas,
Austin.
Appendix.
Taxa, vouchers, localities, and GenBank accession numbers for all sequences analyzed.
Taxo n, origin, voucher, GenBank accession numbers: (trnL-F; ITS). Missing data: –.
Astraea lobata (L.) Klotzsch, BRAZIL, Bahia, Van Ee 486 (WIS), (EU586999; EU586945); Brasiliocroton mamoninha P.E. Berry & I. Cordeiro, BRAZIL,
Espirito Santo, Pirani 4947 (SPF), (EU586998; EU586944); Croton abutiloides Kunth, ECUADOR, Tunguragua, Riina 1391 (WIS), (EU586957; EU586903);
C. arboreus Millsp., MEXICO, Quintana Roo, Van Ee 472 (WIS), (EU497701; EU478029); C. astroites Dryand., PUERTO RICO, Van Ee 537 (WIS), (EU586955;
EU586902); C. beetlei Croizat, BOLIVIA, Santa Cruz, Riina 1512 (WIS), (EF408090; EU586916); C. cajucara Benth., BRAZIL, Caruzo 95 (SP), (EU586968;
EU586913); C. caracasanus Pittier, VENEZUELA, Riina 1288 (VEN), (DQ227557; DQ227525); C. chocoanus Croizat, ECUADOR, Esmeraldas, Riina 1415
(WIS), (EU586995; EU586941); C. cuneatus Klotzsch, VENEZUELA, Berr y 7589 (WIS), Riina 1491 (WIS), (AY794698; EU497735); C. domatifer Riina &
P.E. Ber ry, VEN EZUELA, Ar agua, Riina 1537 (V EN), (GU296103; G U296102); C. echioides Baill., BRAZI L, Santos 795 (HU EFS), (EU586967; E U586907);
C. eichleri ll. Arg., BRAZIL, Rio de Janeiro, Riina 1525 (WIS), (EU587001; EU586949); C. ekmanii Urb., CUBA, Holguín, HAJB 81845 (WIS, HAJB),
(EF408137; EF421760); C. erythroxyloides Baill., BRAZIL, Caruzo 74 (SP), (EU586992; EU586938); C. glandulosepalus Millsp., BELIZE, Vincent 6058
(MU), ( EU478126; EU477888); C. glandulosus L., USA, Wisconsin, Van Ee 512 (WIS), (EU497713; EU478066); C. gossypiifolius Vah l., VE NEZ UE LA, Riina
1261 ( WIS), (AY971301; AY971212); C. helicoideus Müll. Arg. (= C. micans Sw.), BOLIVIA, Wood 140 86 (LPB), (EU586956; EU586902); C. hircinus Ve nt. ,
VENEZUELA, Caracas, Riina 1291 (WIS), (EU478127; EU477889); C. hirtus L’Her., BRAZIL, Lima 345 (SPF), (EU478160; EU478070); C. huberi Steyerm.,
VENEZUELA, Miranda, Berry 7590 (WIS), (AY971305; AY971217); C. lepidotus A.DC., MADAGASCAR, Schatz 3845 (MO), (EU497719; EU497740);
C. malambo Karst., COLOMBIA, Zarucchi 3856 (MO), (AY971315; AY971228); C. megalodendron Müll. Arg., VENEZUELA, Miranda, Riina 1290 (WIS),
(EU586996; EU586942); C. megistocarpus J.A. González & Poveda, COSTA RICA, Jimenez 1527 (MO), (EF408130; EF421787); C. organensis Baill., BRA-
ZIL, Rio de Janeiro, Caruzo 90 (WIS), (EU586969; EU586914); C. pachypodus G.L. Webster; COSTA RICA, San José, Val ver de 10 43 (MO), (EU587004;
EU586953); C. palanostigma Klotzsch; PERU, Loreto, Riina 1492 (WIS), (EU586997; EU586943); C. poecilanthus Urb., PUERTO RICO, Van Ee 551 (WIS),
(EF408122; EF421782); C. priscus Croizat, BRAZIL, Sao Paulo, Riina 1535 (WIS), (EU587002; EU586950); C. sampatik Müll. Arg., PERU, Pasco, Riina
144 7 (WIS), (EF408133; EF421792); C. sapiifolius Müll. Arg., BRAZIL, Bahia, Lima 667 (CEPEC), (EF408150; EF421754); C. schiedeanus Schltdl., COSTA
RICA, Aguilar 886 (MO), (AY971331; AY971246); C. skutchii Standl., COSTA RICA, Cartago, Van Ee 597 (WIS), (EU478166; EU478100); C. smithianus
Croizat, COSTA RICA, Aguilar 2263 (MO), (–; EU478101); C.gnaphalii Baill, ARGENTINA, Belgrano 423 (SI), ( EU586994; EU586940); C. yavitensis
Croizat, BOLIVIA, Beck 5710 (LPB), (EU586973; EU586918); C. gratissimus Burch., ZAMBIA, Songwe Gorge, Zimba 901 (MO), (AY971341; AY971260).
... The genus Croton (Euphorbiaceae family), with approximately 1300 species of herbs, shrubs, and trees, is one of the largest genera of flowering plants and is mainly distributed in both tropical and subtropical regions of Africa, America, and Asia (Berry et al., 2005;Riina et al., 2010). Croton sylvaticus Hochst. ...
Article
The CH2Cl2-MeOH (1:1, v/v) extract of Croton sylvaticus stem bark selectively exhibited good antiplasmodial activity during preliminary screening against the multidrug-resistant (Dd2) and chloroquine sensitive (3D7) strains of Plasmodium falciparum with IC50 values of 10.57 and 18.47 μg/mL, respectively. After fractionation, the fractions exhibited moderate to good activity with IC50 values ranging from 4.31 to 25.49 μg/mL. The purifi�cation of these fractions led to the isolation of six new labdane-type diterpenoids named sylvacrotonins A–F (1–6), alongside seven known compounds (7–13). Their structures were determined based on the analysis of their NMR and MS data. The isolated compounds were assessed in vitro for their antiplasmodial activity against Pf3D7 and PfDd2. The new labdane-type diterpenoids showed moderate antiplasmodial activity against the two P. falciparum strains, with compound 6 being the most active, with IC50 values of 11.26 and 21.80 μg/mL against PfDd2 and Pf3D7, respectively. Cytotoxic assays were performed on African green monkey kidney Vero cells to ensure that the obtained activity was specific to the parasites. Interestingly, all the tested extracts, fractions, and isolated compounds showed acceptable selectivity (SI > 10) against both Plasmodium strains, except fractions CSB-E (SI = 3.93) and CSB-C (SI = 9.22) against Pf3D7. Globally, with resistance indexes (RIs) lower than 1, all the tested samples were found to be more active against multidrug-resistant (PfDd2) than sensitive (Pf3D7) strains. Furthermore, all the isolated compounds exhibited cell cytotoxicity in the normal Vero cell line. The results from this study confirm the use of C. sylvaticus in malaria-endemic countries for its management and suggest that further lead-optimization studies on hit compounds could drive to the identification of potential lead molecules for antimalarial control and eradication.
... The approach we used with Euphorbia could be adopted to study other large clades where species evolve diplochory and show relatively high variations in seed mass and dispersal structure mass (e.g. Croton, Jatropha;Riina, Ee, Wiedenhoeft, Cardozo, & Berry, 2010;Murthy, Chamundeswari, Goverdhen, & Bahadur, 2013). ...
Article
Full-text available
The combination of two dispersal syndromes (diplochory) brings additional benefits to seeds, yet the relative costs and benefits of the two phases are poorly understood. Our goal is to provide the first quantification to test the long‐standing assumption that there are trade‐offs between the two phases in ballistic–ant diplochory. Dispersal investment data were empirically measured for 91 Euphorbiaceae species across different regions of the world. Dispersal distance data of ballochory (seed dispersal by explosion), myrmecochory (seed dispersal by ants) and diplochory was collated from the literature for 210 records (148 species from 44 families). The data were analysed using Model II regression, Kolmogorov–Smirnov test and t test, complemented by phylogenetic comparative approaches. Across diaspores of diplochorous species, the relative investment in myrmecochory increased more than that in ballochory. Fruit coat mass and elaiosome mass covaried in an allometric manner, regardless of seed mass. Diplochorous diaspores were significantly heavier than diaspores dispersed solely by ballochory. Dispersal distances of the two diplochorous phases were independent and comparable to that of sole ballochory or sole myrmecochory. Our results do not support a trade‐off, but a coordinated dispersal strategy between the two diplochorous phases. Large diaspores may evolve diplochory to overcome dispersal difficulties in term of dispersal distance. As the most comprehensive study of ballistic–ant diplochory, our findings advance the understanding of the relative importance of the two phases in diplochory. A plain language summary is available for this article.
Article
Full-text available
Trichomes are key functional traits that protect plants against abiotic stresses, herbivores and pathogens. In addition to their ecological roles, trichomes show high morphological variability in many lineages, providing useful characters for species characterization and taxon delimitation. Here, we focus on trichomes in Croton, a megadiverse genus (c. 1200 spp.) of Euphorbiaceae, in which these structures are ubiquitous and exhibit significant variation in form, location on the plant body, density and function. Owing to their functional and taxonomic relevance and the current need for further standardization of their terminology and classification, we revised previous trichome classifications used in the genus. We also provide new trichome data by expanding taxon sampling and using different methodological approaches in plant anatomy to improve understanding of the structural patterns of the trichome types present in the genus. On the basis of our results and literature review, we propose a new classification for trichomes in Croton based on well-defined parameters; this clarifies existing redundancies, is consistent with anatomical evidence and can be easily updated as new types are revealed in future surveys. This contribution will improve communication among taxonomists and promote more meaningful morphological comparability across systematics studies, including those focusing on morphological evolution and functional traits.
Article
Full-text available
Croton ( Euphorbiaceae ) is a very large genus of over 1200 species of herbs, shrubs and trees occurring mainly in the (sub)tropics. The major diversity of Croton is within the Neotropics; and there are comparatively few species in South-East Asia. This contribution provides a revision of Croton from the Sumatran region in Indonesia, including descriptions and a determination key. In total sixteen species are recognized for Sumatra, including four species new to science, C. beccarii , C. scalaeus , C. simalurensis , and C. viridifolius . Two new synonyms and lectotypification of numerous names are also included. Besides a taxonomic treatment, a molecular analysis, using the nuclear ribosomal internal transcribed spacer (ITS), based on former data extended with Sumatran species, was performed to create a phylogeny that places the Sumatran species in a phylogenetic context. In total 13 new sequences were created, which were combined with 56 sequences obtained from GenBank. The results from the phylogenetic analysis, based on Bayesian inference, are in line with the phylogeny as published for mainly the Australian species of Croton by Van Ee et al. (2015).
Article
Full-text available
Micronesian islands taxa exhibit high endemism rates but very little is known about their biogeographical histories. The lack of systematic biogeography is mainly due to insufficient phylogenetic research in Micronesia. With the recent rise of molecular biogeographic data published, we were able to, for the first time, answer fundamental biogeography questions by reviewing and analyzing numerous geological, ecological, and evolutionary studies. This review, in addition to providing an overview of Micronesian geological history, confirmed the importance of long‐distance dispersal mechanisms and founder‐event speciation, and morphological and physiological adaptations of plant propagules to cross vast stretches of ocean by wind, ocean currents, bird, or bat dispersal. These adaptations to habitat and geological features including reef types, determined colonization success as well as inland dispersal and speciation mechanisms. We further identified the source areas of the Micronesian biota and reconstructed historical dispersal scenarios: a dominant Austro‐Melanesian dispersal scenario, an Indo‐Malaysian connecting to the Austro‐Melanesian dispersal scenario, and a Neotropical American and an African dispersal scenario towards Micronesia. Most generic origins were estimated between the Eocene and Miocene and dispersed to Micronesia between the Miocene and Pleistocene. This article is protected by copyright. All rights reserved.
Article
Full-text available
Identifying traits contributing to the success of organisms growing in harsh environments has always been of great interest in evolutionary ecology. Here we focused on morphological features that could be advantageous for the survival of Croton mollis, a slender shrub occurring in the extreme white-sand habitats of the Amazonian igapos (black/clear-water rivers). We examined macro-morphological characters, as well as leaf micro-morphology using light microscopy and scanning electron microscopy. We also estimated areas of habitat suitability for C. mollis based on climatic, soil, and elevation variables using species distribution modeling (SDM) and available herbarium collections. Our results show that C. mollis presents morphological attributes similar to those of rheophytes, which may allow the species to overcome drastic seasonal changes in water level. We report the presence of five secretory structures (extrafloral nectaries, colleters, idioblasts, glandular trichomes, and laticifers) in the leaves. The size and position of the secretory idioblasts support the hypothesis of the existence of a transitional storage for secondary metabolites from idioblasts to secretory trichomes. Besides glandular trichomes, we also found stellate trichomes and trichome-like emergences in leaves. We hypothesize that these emergences could have an important role in the species’ survival due to their capacity of absorbing atmospheric moisture. Finally, the SDM supported the known habitat preferences of C. mollis and estimated a relatively wider geographic range than the currently known distribution based on herbarium records, suggesting that collecting efforts in the Amazonian region need to be increased in the future.
Article
Full-text available
Croton is one of the largest angiosperm genera, with > 1200 species in the tropics worldwide. The arborescent Croton section Cleodora stands out for its disjunct geographical distribution with two main centres of diversity in the Amazonian and Atlantic Forest regions, separated by the Caatinga, Cerrado and Chaco biomes (the ‘Dry Diagonal’). This disjunction is found in other Neotropical lineages and attributed to Neogene geological and climatic events. We inferred a nearly complete phylogenetic reconstruction of Croton section Cleodora based on DNA sequences of nuclear ITS and five plastid regions (rps16, trnH-psbA, trnL-F, trnT-L and ycf1). We further estimated divergence times and reconstructed ancestral ranges using Bayesian methods. Our results show that Croton section Cleodora is monophyletic with two main clades; we also confirm the phylogenetic adscription of eight Croton spp. recently described or assigned to section Cleodora based on morphology. Divergence of section Cleodora from its sister clade occurred c. 25 Mya, and diversification within this group started 20 Mya. Biogeographic analyses suggest the section originated in the Amazonian region, from where it dispersed to other forested Neotropical regions, including the Atlantic Forest. Divergence between Amazonian and Atlantic Forest taxa appears to have been triggered by the formation of the South American Dry Diagonal.
Article
Full-text available
The present study aimed to morphologically characterize the pollen grains of species of Brasiliocroton mamoninha and B. muricatus in order to expand palynological knowledge of the group and provide additional micromorphological information useful for studies of the relationships within Crotoneae. Polleniferous material was acetolysed and described using light, scanning electron and transmission microscopy. The pollen grains of the two species of Brasiliocroton were apolar, spherical, inaperaturate and medium-sized with a Croton ornamentation pattern. The Croton pattern of B. mamoninha is formed by subtriangular pila, >5-plicate with pointed apices, while for B. muricatus it is formed by subcircular to circular pila, <5-plicate with rounded apices. Ultrastructural analysis revealed a stratified sexine in both species, whereas variation in the pattern of deposition of the foot layer allowed the species to be palynologically distinguished from each other. The composition of the Croton pattern of ornamentation, the morphology of the ultrastructure of the wall and the range of variation in the diameter of the pollen grains differed between the studied species, and thus are diagnostic characters. Pollen morphology allowed the two species of Brasiliocroton to be palynologically separated, and provided data that will be useful for taxonomically circumscribing these taxa.
Article
Full-text available
Background and aims: Wood is a major innovation of land plants, and is usually a central component of the body plan for two major plant habits: shrubs and trees. Wood anatomical syndromes vary between shrubs and trees, but no prior work has explicitly evaluated the contingent evolution of wood anatomical diversity in the context of these plant habits. Methods: Phylogenetic comparative methods were used to test for contingent evolution of habit, habitat and wood anatomy in the mega-diverse genus Croton (Euphorbiaceae), across the largest and most complete molecular phylogeny of the genus to date. Key results: Plant habit and habitat are highly correlated, but most wood anatomical features correlate more strongly with habit. The ancestral Croton was reconstructed as a tree, the wood of which is inferred to have absent or indistinct growth rings, confluent-like axial parenchyma, procumbent ray cells and disjunctive ray parenchyma cell walls. The taxa sampled showed multiple independent origins of the shrub habit in Croton, and this habit shift is contingent on several wood anatomical features (e.g. similar vessel-ray pits, thick fibre walls, perforated ray cells). The only wood anatomical trait correlated with habitat and not habit was the presence of helical thickenings in the vessel elements of mesic Croton CONCLUSIONS: Plant functional traits, individually or in suites, are responses to multiple and often confounding contexts in evolution. By establishing an explicit contingent evolutionary framework, the interplay between habit, habitat and wood anatomical diversity was dissected in the genus Croton Both habit and habitat influence the evolution of wood anatomical characters, and conversely, the wood anatomy of lineages can affect shifts in plant habit and habitat. This study hypothesizes novel putatively functional trait associations in woody plant structure that could be further tested in a variety of other taxa.
Article
Background and aims Although trichomes are typical features of the species-rich genus Croton, little is known about their development, anatomical structure and function. This study aims to characterize the anatomy of leaf trichomes of three Croton species restricted to Brazilian rock outcrops, and investigate their functional role in an environment with limited soil water availability. Methods Samples of leaves at different stages of maturity were submitted to standard anatomical light and scanning electron microscopy techniques to perform a structural and developmental study. Atmospheric water uptake was investigated through the absorption of aqueous solutions of berberine hemisulfate and Indian ink. Key results The so-called trichomes actually correspond to emergences because they originate from protodermal and ground meristem activity rather than from epidermal tissue. These emergences show a complex anatomy, are closely associated with the mesophyll sclereids and might be involved in leaf atmospheric water uptake. The outermost cells of the emergences are arranged radially and present non-lignified thick walls. We demonstrate that an aqueous solution can effectively penetrate through the leaf emergences, continuing through sclereids and reaching the vascular bundle or flowing across the mesophyll maximizing water distribution in the leaf tissue. Conclusions The complex structure and hypothesized function of Croton trichome-like emergences represent a novel finding in Euphorbiaceae. In the three species studied, these leaf emergences might play a role in atmospheric water absorption and they may be one of the clues for the occurrence of Croton in habitats with limited soil water supply where this genus is usually species-rich and abundant.
Article
Full-text available
The phylogenetic relationships of Croton section Cyclostigma sensu Webster were examined using nuclear ribosomal ITS and plastid spacer trnL-F sequences from 48 of the 63 species in the group, including members of the four recognized subsections and representing the group's full morphological and geographic range. The two gene phylogenies concurred that section Cyclostigma sensu Webster is polyphyletic, with its members appearing in nine different clades throughout Croton. The clade that retains the taxonomic type of the section and about half of the species sampled is redefined here as a reduced, monophyletic section Cyclostigma, and the remaining species are placed in other sections or informal clades. The ITS and chloroplast phylogenies are discordant regarding the position of the redefined section Cyclostigma suggesting that reticulate evolution might have played a role in the evolutionary history of this group. These results highlight that the tree habit has evolved on multiple occasions in Croton, and that there are other morphological characters that can effectively distinguish different tree lineages in the genus.
Article
Full-text available
Absolute criteria for evaluating cladistic analyses are useful, not only because cladistic algorithms impose structure, but also because applications of cladistic results demand some assessment of the degree of corroboration of the cladogram. Here, a means of quantitative evaluation is presented based on tree length. The length of the most-parsimonious tree reflects the degree to which the observed characters co-vary such that a single tree topology can explain shared character states among the taxa. This "cladistic covariation" can be quantified by comparing the length of the most parsimonious tree for the observed data set to that found for data sets with random covariation of characters. A random data set is defined as one in which the original number of characters and their character states are maintained, but for each character, the states are randomly reassigned to the taxa. The cladistic permutation tail probability, PTP, is defined as the estimate of the proportion of times that a tree can be found as short or shorter than the original tree. Significant cladistic covariation exists if the PTP is less than a prescribed value, for example, 0.05. In case studies based on molecular and morphological data sets, application of the PTP shows that: 1. (1) In the comparison of four different molecular data sets for orders of mammals, the sequence data set for alpha hemoglobin does not have significant cladistic covariation, while that for alpha crystallin is highly significant. However, when each data set was reduced to the 11 common taxa in order to standardize comparison, reduced levels of cladistic covariation, with no clear superiority of the alpha crystallin data, were found. Morphological data for these 11 taxa had a highly significant PTP, producing a tree roughly congruent with those for the three molecular sets with marginal or significant PTP values. Merging of all data sets, with the exclusion of the poorly structured alpha hemoglobin data, produced a data set with a significant PTP, and provides an estimate of the phylogenetic relationships among these 11 orders of mammals. 2. (2) In an analysis of lactalbumin and lysozyme DNA sequence data for four taxa, purine-pyrimidine coding yields a data set with significant cladistic covariation, while other codings fail. The data for codon position 3 taken alone exhibit the strongest cladistic covariation. 3. (3) A data set based on flavonoids in taxa of Polygonum initially yields a significant PTP; however, deletion of identically scored taxa leaves no significant cladistic covariation. 4. (4) For mitochondrial DNA data on population genome types for four species of the crested newt, there is significant cladistic covariation for the set of all genome types, and among the five mtDNA genome types within one of the species. However, a conditional PTP test that assumes species monophyly shows that no significant cladistic covariation exists among the fur species for these data. 5. (5) In an application of the test to a group of freshwater insects, as preliminary to biological monitoring, individual subsets of the taxonomic data representing larval, pupal, and adult stages had non-significant PTPs, while the complete data set showed significant cladistic structure.
Article
Maximum fruiting of the trees coincides with the period of inundation, and the specific timing of diaspore release appears related to special dispersal mechanisms, dormancy, and/or requirements for germination and seedling establishment. Due to the possession of specific tissues or other devices that provide buoyancy, the diaspores of most tree species of the inundated forests are capable of floating for prolonged periods. A comparison of diaspore characteristics between hydrochorous tree species and their congeners in noninundated habitats reveals recurrent patterns: some lineages that in noninundated habitats possess dehiscent fruits with anemochorous or zoochorous seeds, in inundatable habitats switch to the production of hydrochorous seeds. Other lineages that in noninundated habitats have dehiscent fruits, in inundatable habitats switch to the production of indehiscent hydrochorous fruits. Lineages that in noninundated habitats produce indehiscent fruits remain indehiscent when switching to hydrochory in inundatable habitats. In the inundated forests virtually all diaspores that fall into the water are consumed by fish, with rates of destruction differing greatly. Most hydrochorous diaspores can be dispersed by fish, if they are not destroyed by them (facultative ichthyochory). Others depend on fish for dispersal (obligatory ichthyochory), for example, because their diaspores are very heavy and would sink to the ground under the parent tree, or because their seeds are enclosed in a hard shell from which they can be freed only by the jaws of characins. The dormancy of many kinds of seeds is probably broken by their exposure to the hypoxic conditions that prevail in still water. -from Authors
Article
Collections made in the last two decades have added 9 genera and 42 species to the 35 genera and 100 species included in the Flora of Panama treatment of the Euphorbiaceae published in 1968. The new taxa and combinations proposed in this paper are Richeria dressleri Webster, Phyllanthus gentryi Webster, Tragia correae Huft, Dalechampia canescens Kunth subsp. friedrichsthalii (Muell. Arg.) Webster & Huft, Tetrorchidium costaricense Huft, Tetrorchidium microphyllum Huft, Croton pachypodus Webster, Croton speciosus Muell. Arg. subsp. tacarcunensis Webster, Croton draco Cham. & Schldl. subsp. panamensis (Klotzsch) Webster, Croton billbergianus Muell. Arg. subsp. pyramidalis (J. D. Smith) Webster, Croton santaritensis Huft, Sebastiania panamensis Webster, Gymnanthes dressleri Webster, and Gymnanthes farinosa (Griseb.) Webster. In addition, new or updated keys are provided where appropriate, as well as descriptions and specimen citations.
Article
The estimation procedure utilizes a compatibility analysis between enzyme data sets of the most parsimonious trees constructed from the restriction enzyme. Next, a non-parametric test is given for comparing alternative phylogenies. A 2nd non-parametric test is developed for testing the molecular clock hypothesis. To illustrate the power of these procedures, data derived from the mitochondrial DNA and globin DNA of man and the apes are analyzed. Although previous analyses of these data led to the speculation that 10 times more information would be required to resolve the evolutionary relationships between man with chimps and gorillas, this algorithm resolved these relationships at the 5% level of significance. The molecular clock hypothesis was rejected at the 1% level. The implications of this phylogenetic inference when coupled with other types of data lead to the conclusion that knuckle-walking - not bipedalism - is the evolutionary novelty in mode of locomotion in the primates and that many other hominid features are primitive whereas their African ape counterparts are derived.-from Author