Reprinted from PCAS, ser.

4, 60(May):53-68

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

Series 4, Volume 60, No. 5, pp. 5368, 8 figs., 2 appendices., 2 tables

May 7, 2009

Molecular Phylogeny of Nelsonioideae (Acanthaceae)

and Phylogeography of Elytraria
Rebecca C. Wenk1,2 and Thomas F. Daniel1,2
1Department

of Botany, California Academy of Sciences, 55 Music Concourse Drive, Golden Gate Park,
San Francisco, CA 94118 U.S.A.; Email: [email protected]; 2Department of Biology, San Francisco
State University, 1600 Holloway Avenue, San Francisco, CA 94132 U.S.A.
Nelsonioideae comprise the basal lineage among clades of Acanthaceae. Species in
this subfamily are characterized morphologically by several symplesiomorphies and
at least one synapomorphy (descending-cochlear aestivation pattern of the corolla
lobes). Phylogenetic relationships were studied among five of the seven genera in the
subfamily: Anisosepalum, Elytraria, Nelsonia, Ophiorrhiziphyllon, and Staurogyne.
DNA sequence data from two genic regions (chloroplast trnS-G and nrITS) of 22 collections were used to generate phylogenetic trees using parsimony, maximum likelihood, and Bayesian methods of cladistic analysis. Combined molecular analyses all
strongly support Elytraria as monophyletic, recover a clade consisting of the morphologically similar genera ((Ophiorrhiziphyllon + Staurogyne) Anisosepalum), and
are ambiguous with respect to the topologic position of Nelsonia. Nelsonioideae
include about 170 species that occur primarily in tropical regions of both the Old
World and the New World. Phylogeographic analyses of 11 species of the widely distributed Elytraria using Fitch optimization and dispersal-vicariance analysis (DIVA)
support an African origin for the genus with a single dispersal event to the New
World (likely by rafting of seeds or entire plants) followed by radiation and speciation there.

Acanthaceae contain about 220 genera (Scotland & Vollesen 2000) and some 4,000 species
(McDade et al. 2008) that occur primarily in tropical and subtropical regions worldwide. Four morphologically distinctive subfamilies were recognized by Lindau (1895): Nelsonioideae, Mendoncioideae, Thunbergioideae, and Acanthoideae. Recent molecular studies have confirmed monophyly for each of these, but suggest that Mendoncioideae and Thunbergioideae might be better treated
as sister taxa of a single subfamily, Thunbergioideae (McDade et al. 2008; Borg et al. 2008). Phylogenetic relationships based on a combination of nuclear and chloroplast DNA sequence data have
been studied in numerous genera in each of these subfamilies except Nelsonioideae (e.g., Hedrn
et al. 1995; Scotland et al. 1995; McDade et al. 2000, 2005, 2008; Manktelow et al. 2001; Kiel et
al. 2006; Daniel et al. 2008; Borg et al. 2008). Based on a limited sampling of genera, Nelsonioideae have been shown to be basal to all other lineages of Acanthaceae (e.g., Scotland et al.
1995; McDade et al. 2000, 2008). Studies of phylogenetic relationships and biogeographic patterns
based on phylogenetic trees (phylogeography) among genera of this subfamily are particularly
desirable to better understand aspects of morphological and molecular evolution in the family and
among related lineages of Lamiales.
Nelsonioideae include about 170 species in seven genera that occur primarily in tropical
53

54

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

Series 4, Volume 60, No. 5
TABLE 1. Genera, species richness, and geographic distribution of genera of Nelsonioideae. Total
number of species per genus is followed in parentheses by number sampled here.
Genera

No. Species (no. sampled)

Geographic Distribution

Anisosepalum E. Hossain

3 (1)

tropical Africa

Elytraria Michx.

22 (11)

tropical Africa, Madagascar, Indian subcontinent,

New World
northern South America

Gynocraterium Bremek.

1 (0)

Nelsonia R. Br.

15 (1)

Ophiorrhiziphyllon Kurz

1 (1)

tropical Africa, southern Asia, Australia, New

World (native?)
southeastern Asia

Saintpauliopsis Staner

1 (0)

tropical Africa

Staurogyne Wall.

~140 (1)

Africa, Asia, northern Australia, New World

regions of both the Old World (OW) and the New World (NW) (Table 1). The subfamily is characterized by the following morphological traits: lack of cystoliths (symplesiomorphy), lack of retinacula (symplesiomorphy), relatively large numbers of ovules (symplesiomorphy), and descending-cochlear aestivation pattern of the corolla lobes (synapomorphy). Although usually treated
within Acanthaceae, this assemblage of genera has also been treated as a tribe of Scrophulariaceae
(Bremekamp 1953, 1965) and as a distinct family (Sreemadhavan 1977; Lu 1990).
In this study, we present putative phylogenetic relationships among five of the seven genera of
nelsonioids based on sequence data from the internal transcribed spacer region of the nuclear ribosomal repeat (nrITS) and the non-coding chloroplast trnS-G intergenic spacer. In addition, sampling within the second largest and most widely distributed genus of the subfamily, Elytraria, was
more extensive so that phylogenetic and biogeographic relationships of species in this genus might
begin to be assessed. Specifically, our goals include: 1) proposing generic relationships within the
subfamily, 2) testing monophyly for this sampling of Elytraria, 3) proposing specific relationships
within Elytraria, 4) establishing an area of origin for Elytraria, and 5) providing additional biogeographic insights based on phylogeographic analyses.

MATERIALS AND METHODS

TAXON SAMPLING. Full or partial sequence data of 22 collections, representing five of the
seven genera of Nelsonioideae and 11 of the 22 species of Elytraria, were obtained (Appendix 1).
We were unable to obtain sequence data for samples of Gynocraterium, Saintpauliopsis, and
E. tuberosa; samples of the other eight species of Elytraria were unavailable. Indeed, many species
of Elytraria remain poorly collected due to their limited distributions. For example, E. klugii is
known only by the type collection from Peru. Because one of our goals was to elucidate major biogeographic relationships in Elytraria, and because several species are poorly known or collected,
our taxon sampling for phylogeographic analyses sometimes included only one representative from
a geographic area (e.g., only one of the six species endemic to Cuba was sampled). Aphelandra
maculata (Acanthoideae) was used as an outgroup; its sequence data were obtained from GenBank.
Appendix 1 lists voucher data for each taxon sequenced including: taxon name, GenBank accession number, source, collection data, and type of material (herbarium specimen or fresh sample
dried in silica).
MOLECULAR METHODS. Total genomic DNA was extracted from herbarium samples up to
80 years old or from fresh plant material dried in silica gel by using DNEasy kits (Qiagen Inc.,

WENK & DANIEL: MOLECULAR PHYLOGENY OF NELSONIOIDEAE

55

Valencia, CA) following the manufacturers instructions. The nrITS and trnS-G regions were chosen for sequencing because of previous success and significant phylogenetic signal for those loci
in other species of Acanthaceae (McDade et al. 2005).
DNA was amplified in 25-l reactions. The nrITS region was amplified by using a combination of the universal primers its4 (5-TCC TCC GCT TAT TGA TAT GC-3) and its5 (5-GGA AGT
AAA AGT CGT AAC AAG G-3) (White et al. 1990) and the plant-specific primers C26A (5-GTT
TCT TTT CGT CCG CT-3) and N-nc18S10 (5-AGG AGA AGT CGT AAC AAG-3) (Wen &
Zimmer 1996). The its4, its5 PCR reaction contained 2.5 l 10X buffer, 0.438l 10 mM dNTPs,
0.625 l 50 mM MgCl2, 0.25 units of Biolase Taq polymerase (Bioline, Randolph, MA), and 0.54
l of each primer (25mM). The PCR program consisted of a denaturing step for 6 minutes at 94C,
35 amplification cycles of 60s at 94C, 60s at 5056C, 80s at 72C, and a final extension step of
6 minutes at 72C in a Biorad MyCycler thermal cycler (Biorad, Hercules, CA). The C26A,
N-nc18S10 PCR reaction contained 2.53.0 l 10X buffer, 0.4381.0 l 10 mM dNTPs,
0.6250.75 l 50 mM MgCl2, 0.25 units of Biolase Taq polymerase, and 0.54 l of each primer
(25mM). The PCR program consisted of a denaturing step for 10 minutes at 94C, 35 amplification cycles of 60s at 94C, 6080s at 50C, 80s at 72C, and a final extension step of 10 minutes
at 72C in a Biorad MyCycler thermal cycler. The trnS-G spacer was amplified by using primers
trnS-GCU (5-GCC GCT TTA GTC CAC TCA GC-3) and trnG-UCC (5-GAA CGA ATC ACA
CTT TTA CCA C) (Hamilton 1999). The trnS-GCU, trnG-UCC PCR reaction contained 2.5 l 10X
buffer, 0.438 l 10 mM dNTPs, 0.625 l 50 mM MgCl2, 0.25 units of Biolase Taq polymerase, and
0.54 l of each primer (25mM). The PCR program consisted of a denaturing step for 4 minutes at
94C, 35 amplification cycles of 45s at 94C, 60s at 57C, 80s at 72C, and a final extension step
of 6 minutes at 72C in a Biorad MyCycler thermal cycler.
PCR products were verified prior to sequencing by using agarose gel electrophoresis; they
were cleaned with ExoSAP (Exonuclease I and Shrimp Alkaline Phosphatase). The PCR product
was prepared for sequencing by using the primers from PCR amplification and BigDye fluorescent
dye-terminator reagent mix (Perkin-Elmer, Foster City, CA). Sequence reaction products were
purified with Sephadex G-50 columns (GE Healthcare, Piscataway, NJ) to remove unincorporated
dye terminators. Sequences were generated on an ABI 3100 automated sequencer (Applied Biosystems, Inc., Foster City, CA) in the Osher Molecular Lab at the Center for Comparative Genomics
at the California Academy of Sciences.
SEQUENCE ALIGNMENT. Sequences were manually edited in Sequencher 4.7 (Gene Codes
Corp., Inc., Ann Arbor, MI). Rough alignments were performed in Sequencher 4.7 and ClustalX
1.83 (Thompson et al. 1997) and sequences were further aligned by eye in MacClade 4.08 (Maddison & Maddison 2005). The nrITS matrix was trimmed to include only the two spacers, ITS1 and
ITS2, and the 5.8S RNA region. Sequences of nrITS could not be obtained for seven (about onethird) of the taxa, possibly due to the low quality of DNA in some of the herbarium specimens. The
trnS-G sequence data for one taxon were not obtained, likely for the same reason.
DATA CONGRUENCE. Congruence of the two data sets (nrITS and trnS-G) was tested using
the partition homogeneity test (Farris et al. 1994, 1995) as implemented by PAUP* (Phylogenetic
Analysis Using Parsimony) version 4.0b10 (Swofford 2002). To test for congruence, the two data
sets were reduced to 14 taxa, i.e., they included only those taxa for which sequences of both regions
were obtained. Parameters were set to 100 replicates and 5,000 maximum trees. The test found the
two data sets to be congruent (p = 0.12), and further phylogenetic analyses were conducted for each
region separately and in a combined data set.
PHYLOGENETIC ANALYSES. Maximum parsimony (MP), maximum likelihood (ML), and
Bayesian inference tests were conducted to create phylogenetic trees. Clade support was calculat-

56

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

Series 4, Volume 60, No. 5

ed by using bootstrap methods (Felsenstein 1985) for MP trees and Bayesian Posterior Probabilities (BPP) for ML and Bayesian trees.
MP analyses were performed in PAUP* by using branch and bound searches for single regions
and a heuristic search for the combined data set. Characters were equally weighted and unordered,
with gaps treated as missing data. Heuristic searches were run using 5,000 random replicates and
TBR branch swapping. Bootstrap analyses for single region data sets were run using 1,000 replicates, TBR branch swapping, 10,000 max trees (5,000 in the combined analysis), and 10 random
addition sequences per bootstrap replicate. Clades with a bootstrap value of > 90% were considered to have strong support, 8090% moderate support, and < 80% weak support.
For the ML and Bayesian analyses, we used MrModeltest 2.2 (Nylander 2004) to find the bestfit model of nucleotide substitution. MrModeltest uses four different hierarchies to calculate the
likelihood ratio between models (Posada & Crandall 2001). We then conducted the ML analyses in
PAUP*, starting with a neighbor joining tree and using the parameters in the best-fit model chosen
by the hierarchical likelihood test. Bayesian analyses were performed in MrBayes 3.1.2 (Huelsenbeck & Ronquist 2001; Ronquist & Huelsenbeck 2003). We used a separate best-fit model to represent each DNA region, then combined the two data sets, and implemented both models by partitioning the data. Two Markov chain Monte Carlo (MCMC) runs were performed, each with one
cold chain and three heated chains for 5,000,000 generations, saving a tree every 1,000 generations.
Likelihood scores were determined to be stationary at 345,000 generations (the burn-in point),
and data prior to this point were discarded. Clades with Bayesian Posterior Probabilities > 95%
were considered to have strong support, 9095% moderate support, and < 90% weak support.
CONSTRAINT ANALYSIS. An alternative phylogenetic hypothesis was evaluated using MacClade to create a tree that constrained the clade of interest. This tree was loaded into PAUP* as a
constraint tree, where PAUP* used the maximum parsimony search criteria described above to find
the shortest tree consistent with the new tree topology. The difference in length between the original and constrained trees provides an indication of the parsimony cost (additional steps) involved
in accepting the alternative hypothesis.
PHYLOGEOGRAPHIC ANALYSES. Ancestral areas for Elytraria were located using both Fitch
optimization (Maddison et al. 1992) and dispersal-vicariance analysis (DIVA 1.1; Ronquist 1996,
1997). Fitch optimization is run in PAUP* and codes geographic area as a single multistate character that is treated as unordered. DIVA codes for presence/absence and uses each geographic area
as a separate character. DIVA requires a fully resolved tree. For both of these analyses, a pruned
version of the ML phylogeny shown in Figure 6 was used. To simplify the dataset, we included a
single branch per species. Each taxon used in the analysis (with the exception of E. imbricata) was
assigned to one of six geographic areas: Africa, Madagascar, Greater Antilles, southeastern USA,
South America, and western USA/Mexico/Central America. These areas were chosen because most
species of Elytraria are restricted to just one of them. Elytraria imbricata is the only species present in two areas, western USA/Mexico/Central and South America. Fitch optimization assumes dispersal as the only mode of speciation. DIVA finds the optimal ancestral area by assuming vicariance and penalizing for dispersal and extinction events. In DIVA, vicariance and duplication events
carry a cost of zero, whereas dispersal and extinction events each carry a cost of one. Optimizations try to minimize the cost. For the DIVA analysis, we constrained the number of ancestral areas
to include no more than two using the maxareas option. The number of maxareas was chosen on
the basis of the maximum number of areas occupied by an extant species (Mansion & Struwe 2004;
Oberprieler 2005).

WENK & DANIEL: MOLECULAR PHYLOGENY OF NELSONIOIDEAE

57

RESULTS
SEQUENCE DATA. Table 2 summarizes the sequence data characteristics for each genic region
as well as the combined data set, and includes number of OTUs sequenced, aligned length, % GC
content base frequencies, proportion of variable sites, proportion of parsimony informative sites,
range of pairwise distances (for all taxa and for Elytraria only), MP analysis tree parameters, and
each of the models chosen by MrModeltest and used in the ML and Bayesian analyses.
PHYLOGENETIC ANALYSES. MP analysis of the trnS-G data set resulted in two equally most
parsimonious trees. A strict consensus of these trees is moderately well resolved (Fig. 1). There is
moderate to strong support for six clades: Elytraria (Bootstrap Support, BS = 97%), E. imbricata
(BS = 88%), E. imbricata + E. mexicana (BS = 95%), E. bromoides (BS = 90%), E. nodosa (Madagascar) + E. acaulis (Malawi) (BS = 90%), and Anisosepalum + Staurogyne (BS = 98%). The relationships of other species within Elytraria and the relationships of the other genera of Nelsonioideae are poorly supported or are represented by polytomies. The trnS-G ML analysis yielded a
similar tree (Fig. 2). The ML tree shows strong support for monophyly of Elytraria (BPP = 97%),
TABLE 2. Summary of sequence data characteristics for nrITS region, trnS-G spacer, and combined
dataset. Search method used in maximum parsimony analyses: *branch and bound, **heuristic.
nrITS

trnS-G

combined
dataset

15
612
64.10%

21
704
30.00%

22
1316
45.86%

0.194

0.385

0.296

0.33

0.136

0.226

0.311

0.164

0.232

Number of OTUs
Aligned length
% CG content
Base frequencies

0.165

0.315

0.245

Variable sites (proportion)

340 (0.556)

224 (0.318)

564 (0.429)

Parsimony informative sites (proportion)

163 (0.266)

72 (0.102)

235 (0.179)

Pairwise distances, all taxa (range, %)

0.233.8%

0.023.2%

0.125.5%

Pairwise distances, Elytraria only (range, %)

0.211.4%

0.03.7%

0.19.5%

1*

2*

20**

597

283

900

383

118

512

0.674

0.678

0.668

0.642

0.796

0.688

K80+G

HKY85+G

GTR+G

Maximum parsimony analyses

Number of most parsimonious trees found

Length (all characters)

Length (parsimony informative characters)
Consistency index (parsimony informative characters)
Retention index (parsimony informative characters)
Maximum likelihood analyses
Model chosen by MrModelTest
Bayesian analyses
Model chosen by MrModelTest

HKY85+G

58

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

Series 4, Volume 60, No. 5

E. imbricata (BPP = 100%),

E. imbricata + E. mexicana (BPP
= 100%), E. bromoides (BPP =
100%), a Greater Antilles
clade consisting of species of
Elytraria from Cuba (E. planifolia) and Haiti (E. prolifera) (BPP
= 95%), a Gulf of Mexico
clade consisting of (E. caroliniensis + E. macrophylla) +
Greater Antilles clade (BPP =
100%), E. nodosa (Madagascar)
+ E. acaulis (Malawi) (BPP =
100%), E. nodosa + two eastern
African species (BPP = 97%),
and Anisosepalum + Staurogyne
(BPP = 100%). Strong support
FIGURE 1. Strict consensus of two most parsimonious trees from a trnS-G
for a clade of African/Malagasy branch and bound analysis. Values at nodes indicate bootstrap support > 70%.
+ Gulf of Mexico species of L = 597 (all characters), CI = 0.674, RI = 0.642.
Elytraria (BPP = 97%) was
recovered in this tree, but not in
any other analysis. Bayesian
analysis yielded a tree (not
shown) almost identical to the
ML tree, but with E. marginata
sister to the Gulf of Mexico
clade (BPP = 85%) instead of to
the eastern African species of
Elytraria.
MP analysis of the nrITS
data set resulted in a single, moderately well resolved tree (Fig.
3). There is strong support for
monophyly of Elytraria (BS =
99%), the Greater Antilles
clade (BS = 96%), E. nodosa +
FIGURE 2. Maximum likelihood tree based on a trnS-G analysis using the
E. acaulis (BS = 100%), and K80+G model of evolution. Values at nodes indicate Bayesian posterior probOphiorrhiziphyllon + Staurogyne abilities > 90%.
(BS = 100%). All NW Elytraria
form a clade sister to the eastern African/Malagasy clade (BS = 99%). The ML tree (Fig. 4) shows
strong support for Elytraria (BPP = 100%), E. imbricata + E. mexicana (BPP = 100%), E. nodosa
(Madagascar) + E. acaulis (Malawi) (BPP = 100%), NW Elytraria as sister to the eastern
African/Malagasy clade (BPP = 100%), and Ophiorrhiziphyllon + Staurogyne (BPP = 99%). The
Bayesian (not shown) and ML trees were nearly identical, except that E. bromoides formed an
unsupported clade together with E. planifolia and E. prolifera in the former tree.
MP analysis of the combined nrITS + trnS-G data set resulted in 20 equally most parsimonious
trees. A strict consensus of these trees (Fig. 5) is similar to the nrITS MP tree (Fig. 3), but less well

WENK & DANIEL: MOLECULAR PHYLOGENY OF NELSONIOIDEAE

59

resolved due to multiple collapsed branches. This tree shows

moderate to strong BS support
for eight clades: Elytraria (BS =
99%), E. imbricata + E. mexicana (BS = 98%), E. bromoides
(BS = 87%), the Greater
Antilles clade (BS = 98%),
E. nodosa + E. acaulis (BS =
92%), E. nodosa + two eastern
African species (BS = 80%),
Ophiorrhiziphyllon + Staurogyne
(BS = 100%), and Ophiorrhiziphyllon + Staurogyne +
Anisosepalum (BS = 84%). The
combined ML tree (Fig. 6) is
identical to the Bayesian tree (not
shown), except for the placement
FIGURE 3. Single most parsimonious tree resulting from a nrITS branch and
of Nelsonia, which has no BPP bound analysis. Values at nodes indicate bootstrap support > 70%. L = 283 (all
support. The ML tree shows characters), CI = 0.678, RI = 0.796.
strong BPP support for all eight
clades supported in the MP
analysis: Elytraria (BPP =
100%), E. imbricata + E. mexicana (BPP = 100%), E. bromoides (BPP = 100%), the
Greater Antilles clade (BPP =
100%), E. nodosa + E. acaulis
(BPP = 100%), E. nodosa + two
eastern African species (BPP =
97%), Ophiorrhiziphyllon +
Staurogyne (BPP = 100%), and
(Ophiorrhiziphyllon + Staurogyne) Anisosepalum (BPP = 100%).
BPP values also add strong support for the Gulf of Mexico
clade (BPP = 99%), an eastern
African/Malagasy clade (BPP =
FIGURE 4. Maximum likelihood tree based on a nrITS analysis using the
97%), NW Elytraria (BPP = HKY+G model of evolution. Values at nodes indicate Bayesian posterior prob98%), and NW Elytraria + west- abilities > 90%.
ern/central African Elytraria
(BPP = 99%). The eastern African/Malagasy clade is basal in the genus.
PHYLOGEOGRAPHIC ANALYSES. Fitch parsimony optimization resulted in an optimization
tree (Fig. 7) of five steps (dispersals), with the basal node ancestral area being scored unambiguously as Africa. A single trans-Atlantic dispersal event occurred from Africa to western
USA/Mexico/Central America, and from there dispersals occurred to the southeastern USA, the
Greater Antilles, and South America. There was also a dispersal event from Africa to Madagascar.

60

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

Series 4, Volume 60, No. 5

The DIVA reconstruction (Fig. 8)

yielded two alternative ancestral
areas for the basal node. One scenario identifies an ancestral area
of Africa with a subsequent dispersal
event
to
western
USA/Mexico/Central America.
The other scenario identifies a
vicariance event with an ancestral area of Africa and western
USA/Mexico/Central America.
DIVA identified five dispersal
events, four to five vicariance
events, and five to six duplication events, with the number of
vicariance and duplication events
depending on the ancestral area
of the basal node.

FIGURE 5. Strict consensus of 20 most parsimonious trees from a combined

trnS-G and nrITS heuristic analysis. Values at nodes indicate bootstrap support
> 70%. L = 900 (all characters), CI = 0.668, RI = 0.688.

DISCUSSION
Molecular Phylogenetics
RELATIONSHIPS OF GENERA OF NELSONIOIDEAE. From our analyses of samples from five of
the seven genera of Nelsonioideae, a lineage consisting of Staurogyne, Ophiorrhiziphyllon, and
Anisosepalum is recovered in both combined analyses (e.g., Fig. 6, BPP = 100%). Hossain (1971)
hypothesized that based on morphological characters (e.g., calyx five-partite), these three staurogynoid genera (plus Gynocraterium and Saintpauliopsis, neither
of which were included in our
analyses) are more closely related to each other than to Elytraria
and Nelsonia (both of which
have the calyx four-partite with
the posterior lobe incompletely
fused). Indeed, several species of
Anisosepalum were originally
described in Staurogyne, and
Ophiorrhiziphyllon differs from
Staurogyne in rather minor characters (e.g., smaller connective
between anther thecae, stamens
long-exserted vs. included or
slightly exserted in Staurogyne,
FIGURE 6. Maximum likelihood tree based on a combined trnS-G and
and thecae elongate vs. subglo- nrITS analysis using the GTR+G model of evolution. Values at nodes indicate
bose to globose in Staurogyne). Bayesian posterior probabilities > 90%.

WENK & DANIEL: MOLECULAR PHYLOGENY OF NELSONIOIDEAE

Additional sampling should

reveal whether Staurogyne
should be treated more broadly
so as to include some or all of the
four other genera either known or
hypothesized to be closely related to it.
Hossain (1971) indicated
that Nelsonia and Elytraria were
morphologically distinct from
the staurogynoid genera of
Nelsonioideae. Our analyses usually place Nelsonia separate from
either Elytraria or the staurogynoid genera, but are unable to
establish its topologic position
relative to these other nelsonioid
clades. Morphologically, Nelsonia shares some characters with
Elytraria (e.g., inflorescences of
dense spikes bearing alternate to
spirally arranged bracts), other
characters with the staurogynoid genera (e.g., hooklike projections on the seeds), and some
characters unique to the subfamily (e.g., absence of bracteoles).
Divergent taxonomies of Nelsonia have been proposed with the
number of species varying from
one (Hossain 1984) to five
(Vollesen 1994). Additional sampling will enable testing for
monophyly of that genus, infrageneric relationships, and its
placement within Nelsonioideae.
OF
ELYMONOPHYLY
TRARIA. All
phylogenetic
analyses strongly support monophyly of Elytraria (BS = 9799%, BPP = 97-100%). On the
basis of morphological characters, Elytraria has been recognized for more than 150 years
(Nees 1847) as a cohesive taxon
that occurs on multiple continents in both the OW and the

61

FIGURE 7. Fitch optimization tree showing five dispersal events. Area designations (in parentheses): (A) Mainland Africa, (B) Madagascar, (C) Southeastern USA, (D) Western USA/Mexico/Central America, (E) Greater Antilles,
(F) South America.

FIGURE 8. Ancestral area reconstruction based on DIVA, using a pruned

tree from the ML analysis of the combined dataset. DIVA set to maxareas = 2.
Area designations (in parentheses): (A) Mainland Africa, (B) Madagascar, (C)
Southeastern USA, (D) Western USA/Mexico/Central America, (E) Greater
Antilles, (F) South America. Ancestral areas separated by a slash represent
equally parsimonious reconstructions. Results include 45 vicariance events
(allopatric speciation), 56 duplication events (sympatric speciation), and 5
dispersal events, depending on basal node reconstruction.

62

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

Series 4, Volume 60, No. 5

NW. These characters include: 1) inflorescence peduncles (scapes) covered with alternate to spirally arranged scales (probable synapomophy), 2) four-partite calyx with anterior lobe bifid to bilobed
(symplesiomorphy; shared with Nelsonia), 3) seeds lacking minute hooklike projections (probable
symplesiomorphy; shared with Anisosepalum), and 4) touch-sensitive stigma (probable synapomorphy). Molecular data now confirm the taxonomic delimitation of Elytraria, and show that the
morphological characters coincide with the phylogeny recovered in our analyses.
INFRAGENERIC RELATIONSHIPS OF ELYTRARIA AND MONOPHYLY OF E. IMBRICATA. Elytraria
consists of 22 species that occur in tropical, subtropical, and temperate regions (Appendix 2). Concentrations of species are found in tropical Africa (six species; Dokosi 1979), Cuba (6 species;
Borhidi and Muiz 1977; Dietrich 1982), and Mexico (4 species; Daniel and Acosta C. 2003). Hossain (1971, 1972) divided Elytraria into two subgenera. Subgenus Tetrandra consists of the two
species (E. madagascariensis and E. nodosa) endemic to Madagascar; subgenus Elytraria includes
all of the other taxa. Morphologically, the Malagasy species are unique in Elytraria by having four
(vs. two) fertile stamens and an anther connective with (vs. without) an apical appendage. Comparing morphological and anatomical characters of these two species to others of Elytraria, as well as
to other nelsonioid genera, Hossain (1971, 1972) hypothesized that many of the characteristics of
subgenus Tetrandra are primitive within Elytraria. Specifically, he noted the probable reduction of
stamens from four to two in the genus. Among extant nelsonioids, androecial elements consist of
two fertile stamens and no staminodes (Nelsonia and some Elytraria), two fertile stamens and one
or more staminodes (Ophiorrhiziphyllon, most Elytraria, and some Staurogyne), and four fertile
stamens with zero or one staminode (Anisosepalum, Gynocraterium, Saintpauliopsis, most Staurogyne, and the two Malagasy Elytraria). Combined molecular phylogenetic analyses place the
Malagasy endemic, E. nodosa, in a moderately or strongly supported clade together with two
species from eastern Africa (Fig. 5, BS = 80%; Fig. 6, BPP = 97%), at the base of the genus. On
the basis of this topology, Hossains hypothesis that many characters of subgenus Tetrandra are
primitive (specifically the number of stamens) is not supported. Also, topologies of the trees
obtained in this study do not support recognition of two monophyletic subgenera as treated by Hossain. However, based on morphological characters, the two Malagasy endemics would indeed
appear to represent closely related species worthy of infrageneric recognition.
The combined ML analysis (Fig. 6) reveals that the OW species of Elytraria do not form a
monophyletic group. Indeed, the central to western African species E. marginata is strongly supported as sister to the NW clade of the genus in this analysis (Fig. 6, BPP = 99%). In both MP and
ML analyses, a clade consisting of the eastern African species and E. nodosa from Madagascar are
supported as monophyletic (Fig. 5, BS = 80%; Fig. 6, BPP = 97%) and strongly supported as sister to NW Elytraria + E. marginata (Fig. 5, BS = 99%; Fig. 6, BPP = 100%). Constraining OW
Elytraria to monophyly by altering the placement of E. marginata in our combined MP tree (Fig.
5) requires only two additional steps (L=900 vs. 902). The constraint analysis suggests that monophyly of OW Elytraria cannot be rejected by our data. The taxonomy of OW species has yet to be
fully resolved and our samples do not include any plants from Asia (i.e., E. acaulis from India and
Sri Lanka).
In the combined ML analysis, the NW species of Elytraria form a strongly supported clade
(Fig. 6, BPP = 98%), but this clade is not supported in the MP analyses. The topologic positions of
three oft recovered NW lineages: E. imbricata + E. mexicana, E. bromoides, and the Gulf of Mexico clade, vary in each of the analyses. Moderate to strong support for a monophyletic E. imbricata is shown only in the trnS-G phylogenies (Fig. 1, BS = 88%; Fig. 2, BPP = 100%), which have
E. mexicana as sister to E. imbricata. When the nrITS data is added, a monophyletic E. imbricata
either has no support (Figs. 5, 6), or E. mexicana is placed within E. imbricata (Figs. 3, 4). Ely-

WENK & DANIEL: MOLECULAR PHYLOGENY OF NELSONIOIDEAE

63

traria mexicana is known only from semi-arid regions of central and western Mexico (Daniel &
Acosta C. 2003), within the range of E. imbricata. Morphologically, these species appear to be very
similar; they differ primarily in characters of the bracteal apex (acuminate-spinose in E. mexicana
vs. prominently 3-parted, at least among distal bracts, in E. imbricata), bracteal pubescence (abaxially villous in E. mexicana vs. abaxially usually glabrous in E. imbricata), and color of the corolla limb (whitish with dark purple markings on the lobes of the upper lip in E. mexicana vs. blue
with white and yellow markings on the lower lip or rarely entirely white in E. imbricata). The morphological distinctions and taxonomic status of these two species are currently being reevaluated;
however, our results substantiate a close relationship between them.

Phylogeography
DISTRIBUTIONS OF NELSONIOIDEAE. Genera of Nelsonioideae exhibit diverse patterns of
present-day distribution (Table 1). The largest genus, Staurogyne, has concentrations of species in
South America, Africa, and Asia; a few species also occur in southern North America and northern
Australia. Close relatives of Staurogyne, both known (Anisosepalum and Ophiorrhiziphyllon) and
suspected (Gynocraterium and Saintpauliopsis), are restricted to areas with significant concentrations of species of that genus. Nelsonia is pantropical with native occurrences in Africa, Madagascar, Asia, Australia, and possibly the NW (Hossain 1971; Vollesen 1994). Elytraria appears to be
particularly widespread in both the OW and NW tropics, with several NW species either extending
their ranges into (e.g., E. bromoides, E. imbricata) or endemic to (e.g., E. caroliniensis) temperate
regions.
AREA OF ORIGIN AND DISPERSAL OF ELYTRARIA. DIVA provides two alternative ancestral
areas, either Africa or Africa and western USA/Mexico/Central America. Fitch optimization results
in an ancestral area of Africa, with a single dispersal event to the New World, from western or central Africa across the Atlantic Ocean. Because both analyses include an optimization of Africa at
the ancestral node of Elytraria, most evidence supports Africa as the area of origin for this genus.
Because available evidence and analyses to date (see McDade et al. 2005, 2008) establish that
Acanthaceae evolved subsequent to the breakup of the Gondwanan landmass, dispersal is the most
appropriate explanation for intercontinental disjunctions. Most recent phylogenetic studies on other
clades of Acanthaceae have revealed an OW origin followed by a single dispersal event to the NW
with subsequent diversification there (i.e., Acantheae, McDade et al. 2005; Isoglossinae, Kiel et al.
2006; Tetramerium lineage of Justicieae, Daniel et al. 2008; Thunbergioideae, Borg et al. 2008).
Our results suggest a similar scenario for Elytraria as well.
Renner (2004) studied the mechanisms for plant dispersal across the tropical Atlantic. Wind,
birds, water, and floating rafts of vegetation are all possible mechanisms of trans-oceanic dispersal. Strong east-to-west surface winds rarely make it across the entire central Atlantic, although
Saharan dust does occasionally make it to the New World (Renner 2004). Seeds of Elytraria are
small (between 0.3 and 1.0 mm in diameter), but lack attributes (e.g., wings) that might make them
candidates for wind dispersal. Seeds of Elytraria may have been transported across the ocean by
sticking to mud on the feet of birds. Today there are no avian species that have tropical transAtlantic migration routes (Renner 2004), though a few sea birds are carried from West Africa
across the Atlantic to North America by tropical storms (Rappole et al. 2000). Bird transportation
remains a possible, though unlikely, mode of trans-Atlantic dispersal for Elytraria.
Oceanic surface water currents influence the last two modes of dispersal under consideration:
water and rafting. Todays patterns of oceanic surface currents have changed little since the end of
the Late Cretaceous (Parrish 1993). Currents in the northern South Atlantic are primarily deter-

64

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

Series 4, Volume 60, No. 5

mined by the placement of the continents, and modern day ocean circulation patterns are thought
to have arisen shortly after the formation of the South Atlantic (Parrish 1993). Today, currents are
predominantly east-to-west, with a single west-to-east current that originates well offshore (Renner 2004).
The small size and unprotected coating of seeds of Elytraria render them unlikely to survive
in salt water for prolonged periods of time. A raft of vegetation carrying seeds or entire plants provides the best explanation for the present day distribution of Elytraria in both the OW and the NW.
Equatorial currents can transport large mats of vegetation coming out of the mouths of the Niger
and Congo Rivers across the Atlantic Ocean in less than two weeks (Renner 2004; Measey et al.
2007). There is evidence suggesting that east-to-west trans-Atlantic rafting events may account for
dispersal in multiple plant and animal groups (Carranza et al. 2000; Cronn and Wendel 2004; Vidal
et al. 2008).
While the amphi-Atlantic distribution of Elytraria can be explained by a single oceanic dispersal event, several shorter dispersals must also have taken place to account for its multiple insular occurrences. Two species of Elytraria, E. nodosa and E. madagascariensis, are endemic to the
island of Madagascar. These species share several morphological characters not found in any other
species of Elytraria (see above), and can be assumed to be sister species based on these putative
synapomorphies. All molecular phylogenetic analyses strongly support placing E. nodosa sister to
E. acaulis from eastern Africa (e.g., Fig. 5, BS = 92%; Fig. 6, BPP = 100%), suggesting a single
dispersal event to the island from eastern Africa. Madagascar lies 400 km off the eastern coast of
Africa and has maintained this position relative to the mainland for the last 120130 my (Rabinowitz & Woods 2006). Rabinowitz & Woods (2006) found that the assumption of previous studies supporting the exposure of dry land in the Mozambique Channel, connecting Madagascar to the
mainland, to be incorrect. Rather, they found that it is unlikely there has been any land exposure
since before the middle Eocene (50 mya), when the first mammals migrated to Madagascar. Rabinowitz & Woods (2006) hypothesized that the early mammals probably arrived by swimming or
by floating on vegetation mats, the most likely method for dispersal of Elytraria there as well.
Island dispersal has also occurred in the Gulf of Guinea, where the widespread West African
Elytraria marginata is found on the islands of So Tom and Prncipe. These oceanic islands are
part of the Cameroon Line of volcanoes and lie over 200 km from the mainland. With ages of at
least 13 and 31 million years (Lee et al. 1994), respectively, So Tom and Prncipe have never
been connected to continental Africa (Jones 1994). Measey et al. (2007) propose that the colonization of So Tom by amphibians was via rafting events. Major sea surface currents in the Gulf of
Guinea would carry any rafts of debris from the Niger and Congo rivers directly towards So Tom,
making rafting a plausible explanation for the dispersal of Elytraria to these islands.
In the Caribbean Sea, Elytraria occurs on several islands. In some cases, dispersals have led
to the formation of new species (e.g., Cuba and Hispaniola); in other cases, insular occurrences of
E. imbricata likely result from recent colonizations or persistence of the species on islands previously connected to the American mainland during periods of lower sea levels (e.g., Trinidad, Tobago, Bonaire, Curaao, Isla San Andrs).

Future Studies
These preliminary findings address phylogenetic relationships of several constituent genera of
Nelsonioideae, relationships of species in Elytraria, and biogeographic patterns based on the phylogenies recovered. Relationships of the seven genera of Nelsonioideae remain unresolved due to
weak or no support in the lower branches of the trees and absence of data for some genera

WENK & DANIEL: MOLECULAR PHYLOGENY OF NELSONIOIDEAE

65

(Gynocraterium and Saintpauliopsis). Future studies on Nelsonioideae will include additional field
and herbarium work in order to obtain collections of more taxa (i.e., species of Elytraria and other
genera of Nelsonioideae), and will add sequences from additional genic regions. These additions
should result in more comprehensive and better resolved trees that can further address taxonomic
composition and monophyly of the subfamily, whether staurogynoid genera should be treated in
a broader Staurogyne, the infrageneric classification of Staurogyne proposed by Hossain (1971),
and phylogeography of the entire subfamily. Phylogeography of this basal lineage of Acanthaceae
should be particularly revealing because the distribution patterns of extant genera common to both
the OW and the NW (Elytraria, Nelsonia, and Staurogyne) suggest a more dynamic history of
intercontinental dispersals than is known for other lineages of Acanthaceae studied to date.

ACKNOWLEDGMENTS
This study represents a portion of a Master of Science thesis from San Francisco State University by the first author. Molecular sequencing was supported by the SFSU Biology Department IRA
Research Grant. Wenks travel to meetings was funded by the SFSU Graduate Student Council in
Biology Travel Grants and the Department of Botany at the California Academy of Sciences.
Daniels studies of Elytraria were supported, in part, by a grant from the National Science Foundation (DEB-0743273). We thank A. Carmichael, F. Cipriano, B. Cruz, K. Deiner, M. Flannery, P.
Fritsch, W.B. Simison, and G. Spicer for help with molecular sequencing and phylogenetic analyses. Also, we thank F. Almeda, D. Champluvier, J.L. Clark, and L. McDade for providing DNA
samples. We are grateful to the following herbaria for allowing us to borrow specimens for morphological studies and/or DNA sampling: A, BR, CAS, DS, F, GH, MO, NY, P, S, UC, and US.

LITERATURE CITED
BORG, A.J., L.A. MCDADE, AND J. SCHNENBERGER. 2008. Molecular phylogenetics and morphological evolution of Thunbergioideae (Acanthaceae). Taxon 57:811822.
BORHIDI, A. AND O. MUIZ. 1977. Notas Sobre Acantceas Cubanas I. Oplonia y Elytraria. Acta Botanica
Academiae Scientientiarum Hungaricae 23:303317.
BREMEKAMP, C.E.B. 1953. The delimitation of Acanthaceae. Proceedings of the Koninklijke Nederlandse
Academie van Wetenschappen. Series C: Biological and Medical Sciences 56:533546.
BREMEKAMP, C.E.B. 1965. Delimitation and subdivision of the Acanthaceae. Bulletin of the Botanical Survey
of India 7:2130.
CARRANZA S., E. N. ARNOLD, J. A. MATEO, AND L. F. LOPEZ-JURADO. 2000. Long-distance colonization and
radiation in gekkonid lizards, Tarentola (Reptilia: Gekkonidae), revealed by mitochondrial DNA
sequences. Proceedings pf the Royal Society of London, Series B: Biological Sciences 267:637649.
CRONN, R. AND J.F. WENDEL. 2004. Cryptic trysts, genomic mergers, and plant speciation. New Phytologist
161:133142.
DANIEL, T.F. AND S. ACOSTA C. 2003. Acanthaceae. Pages 1173 in J. Rzedowski and G. Calderon de Rzedowski, eds., Flora del Bajo y de Regiones Adyacentes, Fascculo 117. Instituto de Ecologa, Centro
Regional del Bajo, Ptzcuaro, Mexico.
DANIEL, T.F., L.A. MCDADE, M. MANKTELOW, AND C.A. KIEL. 2008. The Tetramerium lineage (Acanthaceae: Acanthoideae: Justicieae): delimitation and intra-lineage relationships based on cp and nrITS
sequence data. Systematic Botany 33:416436.
DIETRICH, H. 1982. Acanthaceae Cubanae Novae I. Elytraria bissei H. Dietrich, spec. nov. Revista del Jardin
Botnico Nacional (Habana) 3(2):3951.
DOKOSI, O.B. 1979. A new species of Elytraria (Acanthaceae) occurring in East Africa. Adansonia
18:433441.

66

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

Series 4, Volume 60, No. 5

FARRIS, J.S., M. KLLERSJ, A.G. KLUGE, AND C. BULT. 1994. Testing significance of incongruence. Cladistics
10:315319.
FARRIS, J.S., M. KLLERSJ, A.G. KLUGE, AND C. BULT. 1995. Constructing a significance test for incongruence. Systematic Biology 44:570572.
FELSENSTEIN, J. 1985. Confidence limits on phylogenetics: an approach using the bootstrap. Evolution
39:783791.
HAMILTON, M.B. 1999. Four primer pairs for the amplification of chloroplast intergenic regions with intraspecific variation. Molecular Ecology 8:521523.
HEDRN, M., M.W. CHASE, AND R.G. OLMSTEAD. 1995. Relationships in the Acanthaceae and related families
as suggested by cladistic analysis of rbcL nucleotide sequences. Plant Systematics and Evolution
194:93109.
HOLMGREN, P.K., N.H. HOLMGREN, AND L.C. BARNETT. 1990. Index herbariorum, part I: the herbaria of the
world. Regnum Vegetabile 120:1693.
HOSSAIN, A.B.M.E. 1971. Studies in the Classification and Affinities of Acanthaceae. Ph.D. Dissertation, University of Edinburgh, United Kingdom.
HOSSAIN, A.B.M.E. 1972. Studies in Acanthaceae tribe Nelsonieae I: new and re-named taxa. Notes from the
Royal Botanic Garden, Edinburgh 31:377388.
HOSSAIN, A.B.M.E. 1984. Taxonomic notes on the Nelsonia canescens complex (Acanthaceae). Willdenowia
14:397403.
HUELSENBECK, J. P. AND F. RONQUIST. 2001. MRBAYES: Bayesian inference of phylogeny. Bioinformatics
17:754755.
JONES, P.J. 1994. Biodiversity in the Gulf of Guinea: an overview. Biodiversity and Conservation 3:772785.
KIEL, C.A., L.A. MCDADE, T.F. DANIEL, AND D. CHAMPLUVIER. 2006. Phylogenetic delimitation of Isoglossinae (Acanthaceae: Justicieae) and relationships among constituent genera. Taxon 55:683694.
LEE, D.-C., A.N. HALLIDAY, J.G. FITTON, AND G. POLI. 1994. Isotopic variations with distance and time in the
volcanic islands of the Cameroon line: evidence for a mantle plume origin. Earth and Planetary Science
Letters 123:119138.
LINDAU, G. 1895. Acanthaceae. Pages 274353 in A. Engler and K. Prantl, eds., Die natrlichen Planzenfamilien Vol. 4 (3b). Engelmann, Leipzig, Germany.
LU AN-MING. 1990. A preliminary cladistic study of the families of the superorder Lamiiflorae. Botanical
Journal of the Linnean Society 103:3957.
MADDISON, D.R. AND W.P. MADDISON. 2005. MacClade 4: Analysis of Phylogeny and Character Evolution.
Version 4.08. Sinauer, Sunderland, Massachusetts, USA.
MADDISON, D.R., M. RUVOLO, AND D.L. SWOFFORD. 1992. Geographic origins of human mitochondrial DNA:
phylogenetic evidence from control region sequences. Systematic Biology 41:111124.
MANKTELOW, M., L.A. MCDADE, B. OXELMAN, C.A. FURNESS, AND M.-J. BALKWILL. 2001. The enigmatic tribe
Whitfieldieae (Acanthaceae): delimitation and phylogenetic relationships based on molecular and morphological data. Systematic Botany 26:104119.
MANSION, G. AND L. STRUWE. 2004. Generic delimitation and phylogenetic relationships within the subtribe
Chironiinae (Chironieae: Gentianaceae), with special reference to Centaurium: evidence from nrDNA and
cpDNA sequences. Molecular Phylogenetics and Evolution 32:951977.
MCDADE, L.A., T.F. DANIEL, C.A. KIEL, AND K. VOLLESEN. 2005. Phylogenetic relationships among
Acantheae (Acanthaceae): major lineages present contrasting patterns of molecular evolution and morphological differentiation. Systematic Botany 30:834862.
MCDADE, L.A., T.F. DANIEL, AND C.A. KIEL. 2008. Toward a comprehensive understanding of phylogenetic
relationships among lineages of Acanthaceae s.l. (Lamiales). American Journal of Botany 95:11361152.
MCDADE, L.A., S.E. MASTA, M.L. MOODY, AND E. WATERS. 2000. Phylogenetic relationships among Acanthaceae: evidence from two genomes. Systematic Botany 25:106121.
MEASEY, G.J., M. VENCES, R.C. DREWES, Y. CHIARI, M. MELO, AND B. BOURLES. 2007. Freshwater paths across
the ocean: molecular phylogeny of the frog Ptychadena newtoni gives insights into amphibian colonization of oceanic islands. Journal of Biogeography 34:720.
NEES VON ESENBECK, C.G.D. 1847. Acanthaceae. Pages 46519 in A. de Candolle, ed., Prodromus Systematis

WENK & DANIEL: MOLECULAR PHYLOGENY OF NELSONIOIDEAE

67

Naturalis Regni Vegetablis, vol. 11. V. Masson, Paris, France.

NYLANDER, J.A.A. 2004. MrModeltest v2. Program distributed by the author. Evolutionary Biology Centre,
Uppsala University, Sweden.
OBERPRIELER, C. 2005. Temporal and spatial diversification of circum-Mediterranean Compositae-Anthemidae. Taxon 54:951966.
PARRISH, J.T. 1993. The palaeogeography of the opening of the South Atlantic. Pages 827 in W. George and
R. Lavocat, eds., The Africa-South America Connection. Clarendon Press, Oxford, United Kingdom.
POSADA, D. AND K.A. CRANDALL. 2001. Selecting the best-fit model of nucleotide substitution. Systematic
Biology 50:580601
RABINOWITZ, P.D. AND S. WOODS. 2006. The Africa-Madagascar connection and mammalian migrations. Journal of African Earth Sciences 44:270276.
RAPPOLE, J.H., S.R. DERRICKSON, AND Z. HUBLEK. 2000. Migratory birds and the spread of West Nile virus
in the Western Hemisphere. Emerging Infection Diseases 6:319-328.
RENNER, S. 2004. Plant dispersal across the tropical Atlantic by wind and sea currents. International Journal
of Plant Science 165(4 Suppl.):S23S33.
RONQUIST, F. 1996. DIVA version 1.1. Computer program and manual available by anonymous FTP from Uppsala University (ftp.uu.se or ftp.systbot.uu.se).
RONQUIST, F. 1997. Dispersal-vicariance analysis: A new approach to the quantification of historical biogeography. Systematic Biology 46:195203.
RONQUIST, F. AND J.P. HUELSENBECK. 2003. MRBAYES 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19:15721574.
SCOTLAND, R.W. AND K. VOLLESEN. 2000. Classification of Acanthaceae. Kew Bulletin 55:513589.
SCOTLAND, R.W., J.A. SWEERE, P.A. REEVES, AND R.G. OLMSTEAD. 1995. Higher-level systematics of Acanthaceae determined by chloroplast DNA sequences. American Journal of Botany 82:266275.
SREEMADHAVAN, C.P. 1977. Diagnosis of some new taxa and some new combinations in Bignoniales. Phytologia 37:412416.
SWOFFORD, D.L. 2002. PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods). Version 4b10.
Sinauer Associates, Sunderland, Massachusetts, USA.
THOMPSON, J.D., T.J. GIBSON, F. PLEWNIAK, F. JEANMOUGIN, AND D.G. HIGGINS. 1997. The ClustalX windows
interface: flexible strategies for multiple alignment aided by quality analysis tools. Nucleic Acids Research
25:48764882.
VIDAL, N., A. AZVOLINSKY, C. CRUAUD, AND S.B. HEDGES. 2008. Origin of tropical American burrowing reptiles by transatlantic rafting. Biology Letters 4:115118.
VOLLESEN, K. 1994. Taxonomy, ecology and distribution of Nelsonia (Acanthaceae) in Africa. Pages 315325
in J.H. Seyani and A.C. Chikuni, eds., Proceedings of the XIIIth Plenary Meeting AETFAT, Malawi.
WEN, J. AND E.A. ZIMMER. 1996. Phylogeny and biogeography of Panax L. (the ginseng genus, Araliaceae):
inference from ITS sequences on nuclear ribosomal DNA. Molecular Phylogenetics and Evolution
6:167177.
WHITE, T.J., T. BRUNS, S. LEE, AND J. TAYLOR. 1990. Amplification and direct sequencing of fungal ribosomal
RNA genes for phylogenetics. Pages 315322 in M.A. Innis, D.H. Gelfand, J.J. Sninsky, and T.J. White,
eds., PCR Protocols: A Guide to Methods and Applications. Academic Press, Inc., San Diego, California,
USA.

68

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

Series 4, Volume 60, No. 5

Appendix I
Taxa, GenBank accession numbers (trnS-G, nrITS; = sequence not obtained), sources of plant materials from which DNA was extracted for sequencing, and type of material used for sequencing (herbarium
specimen = H vs. silica-dried sample = S). Cultivated plants include native range in parentheses. Abbreviations for herbaria follow Holmgren et al. (1990). Taxa are listed in alphabetical order by genus and species.
Anisosepalum alboviolaceum (R. Ben.) E. Hossain; FJ868921, FJ868907; Champluvier 5,295 (CAS); Mbomo,
Congo; H. Aphelandra maculata (Tafalla ex Nees) Voss; DQ059281, DQ028451, sequences downloaded from GenBank.
Elytraria acaulis Lindau; FJ868940, FJ868920; Goyoler & Paton 3,520 (BR); Malawi; H. Elytraria bromoides Oerst.;
FJ868929, FJ868914; T.F. Daniel 10,310 (CAS); Yucatn, Mexico; S. Elytraria bromoides Oerst.; FJ868930, ; T.F.
Daniel 10,505 (CAS); Yucatn, Mexico; S. Elytraria caroliniensis (J.F. Gmel.) Pers.; FJ868936, FJ868918; 1971292900cv, cultivated at BR (southeastern United States); S. Elytraria imbricata (Vahl) Pers.; FJ868928, FJ868913; T.F. Daniel
10,108 (CAS); Santa Cruz, Bolivia; S. Elytraria imbricata (Vahl) Pers.; FJ868927, ; T.F. Daniel 10,131 (CAS); Santa
Cruz, Bolivia; S. Elytraria imbricata (Vahl) Pers.; FJ868924, FJ868910; T.F. Daniel 10,501 (CAS); Yucatn, Mexico;
S. Elytraria imbricata (Vahl) Pers.; FJ868925, FJ868911; T.F. Daniel s.n. (CAS); Trinidad; S. Elytraria imbricata (Vahl)
Pers.; FJ868923, ; E.J. Lott 5,175 (CAS); Texas, United States; H. Elytraria imbricata (Vahl) Pers.; FJ898626,
FJ868912; Van Proosij et al. 719 (NY); Paradera, Aruba; H. Elytraria macrophylla Leonard; FJ868933, ; Rzedowski
43,302 (CAS); Quertaro, Mexico; H. Elytraria marginata Vahl; FJ868937, ; 19860332cv, cultivated at BR (Democratic Republic of Congo); S. Elytraria mexicana Fryxell & S.D. Koch; FJ868931, FJ868915; T.F. Daniel 5,284 (CAS); Colima, Mexico; H. Elytraria minor O.B. Dokosi; FJ868938, ; Faulkner 1,783A (BR); Tanzania; H. Elytraria nodosa E. Hossain; FJ868939, FJ868919; T.F. Daniel et al. 10,452 (CAS); Antsiranana, Madagascar; S. Elytraria planifolia Leonard;
FJ868934, FJ868916; R.A. Howard et al. 341 (NY); Villa Clara Province, Cuba; H. Elytraria prolifera Leonard; FJ868935,
FJ868917; Plantae Indiae Occidentalis 8,877 (NY); Ile La Gunave, Haiti; H. Nelsonia canescens (Lam.) Spreng.;
FJ868932, ; T.F. Daniel et al. 5,452cv, cultivated at the San Francisco Conservatory of Flowers (Panama); S. Ophiorrhiziphyllon macrobotryum Kurz; , FJ868908; J.F. Maxwell 91-213 (CAS); Mae Hong Song, Thailand; H. Staurogyne
concinnula (Hance) O. Kuntze; FJ868922, FJ868909; Bartholomew & Boufford 6,215 (CAS); Taipei Hsien, Taiwan;
H. Elytraria acaulis Lindau: tropical and southern Africa, Indian subcontinent

Appendix II
Species of Elytraria and their native distributions.
Elytraria bissei H. Dietr.: Cuba
Elytraria bromoides Oerst.: North America, northern Central America
Elytraria caroliniensis (Gmel.) Pers.: southeastern USA
Elytraria cubana Alain: Cuba
Elytraria filicaulis Borhidi & O. Muniz: Cuba
Elytraria imbricata (Vahl) Pers.: North America, Central America, South America
Elytraria ivorensis Dokosi: tropical western Africa
Elytraria klugii Leonard: Peru
Elytraria lyrata Vahl: tropical Africa
Elytraria macrophylla Leonard: Mexico
Elytraria madagascariensis (R. Ben.) E. Hossain: Madagascar
Elytraria marginata Vahl: tropical western and central Africa
Elytraria maritima J.K. Morton: tropical western Africa
Elytraria mexicana Fryxell & S.D. Koch: Mexico
Elytraria minor Dokosi: tropical eastern Africa
Elytraria nodosa E. Hossain: Madagascar
Elytraria planifolia Leonard: Cuba
Elytraria prolifera Leonard: Haiti
Elytraria shaferi (P. Wilson) Leonard: Cuba
Elytraria spathulifolia Borhidi & O. Muniz: Cuba
Elytraria tuberosa Leonard: Ecuador
Copyright 2009 by the California Academy of Sciences
San Francisco, California, U.S.A.