Content uploaded by Siddharthan Surveswaran
Author content
All content in this area was uploaded by Siddharthan Surveswaran on Nov 14, 2017
Content may be subject to copyright.
Research Article
doi: 10.1111/jse.12185
Molecular phylogeny of Glyphochloa (Poaceae,
Panicoideae), an endemic grass genus from the Western
Ghats, India
Kumar Vinod Chhotupuri Gosavi
1,2
*
†
, Shrirang R. Yadav
1
, K. Praveen Karanth
3
, and Siddharthan Surveswaran
3
*
†
1
Department of Botany, Shivaji University, Kolhapur 416004, India
2
Department of Botany, P.S.G.V.P.M. Arts, Science & Commerce College, Shahada, Nandurbar Dist. 425409, Maharashtra, India
3
Centre for Ecological Sciences, Indian Institute of Science, Bangalore 560012, India
†
These authors contributed equally to this work.
*Authors for correspondence. E-mail: kumarvinodgosavi@gmail.com. siddhus@gmail.com
Received 18 January 2015; Accepted 29 August 2015; Article first published online xx Month 2015
Abstract The genus Glyphochloa (Poaceae: Panicoideae: Andropogoneae: Rottboellinae) is endemic to peninsular
India and is distributed on lateritic plateaus of low and high altitude in and around Western Ghats and the Malabar
Coast. The genus presumably originated and diversified in the Western Ghats. Species relationships in the genus
Glyphochloa were deduced here based on molecular phylogenies inferred using nuclear ribosomal ITS sequences
and plastid intergenic spacer regions (atpB-rbcL,trnT-trnL,trnL-trnF), and new observations were made of spikelet
morphology, caryopsis morphology and meiotic chromosome counts. We observed two distinct clades of
Glyphochloa s.l. One of these (‘group I’) includes Ophiuros bombaiensis, and is characterized by a single-awned lower
glume and a base chromosome number of 6; it grows in low elevation coastal areas. The other clade (‘group II’) has a
double-awned lower glume, a base chromosome number of 7, and is restricted to higher elevation lateritic plateaus;
G. ratnagirica may belong to the group II clade, or may be a third distinct lineage in the genus. A sister-group
relationship between group I and II taxa (with or without G. ratnagirica) is not well supported, although the genus is
recovered as monophyletic in shortest trees inferred using ITS or concatenated plastid data. We present a key to
species of Glyphochloa and make a new combination for O. bombaiensis.
Key words: cytology, Glyphochloa, Indian endemic grasses, molecular phylogeny, Ophiuros bombaiensis, spikelet morphology,
Western Ghats.
The exchange of plant species between India and other parts
of Asia is hindered by India’s long peninsular coastline and by
its separation from the rest of Asia by the Himalayan mountain
ranges in the north. Its rich ecosystems have evolved many
endemic taxa (Datta-Roy & Karanth, 2009; Karanth, 2015). The
Western Ghats region of Peninsular India is one of the 34
recognized global biodiversity hotspots (Mittermeier et al.,
2004). India has about 49 endemic genera, including 13 grass
genera (Irwin & Narasimhan, 2011). The majority of these
inhabit the Western Ghats, and the evolution of these
endemic grasses is little studied or understood.
Glyphochloa Clayton (Poaceae) is one of these endemic
grass genera (Irwin & Narasimhan, 2011). It belongs to subtribe
Rottboellinae (subfamily, Panicoideae; tribe, Andropogoneae;
Welker et al., 2014). It is restricted to peninsular India and
comprises 11 species and four varieties (Royal Botanic
Gardens, Kew, 2013). The species of Glyphochloa have been
variously treated under Rottboellia L.f (Hooker, 1894),
Peltophorus Desv., nom. superfl. (Blatter & McCann, 1935)
and Manisuris L. (Bor, 1960; Rao & Hemadri, 1968; Jain &
Deshpande, 1969; Jain & Hemadri, 1969; Jain, 1972; Kulkarni &
Hemadri, 1974). However, it was recognized as a separate
genus by Clayton (1980), and is distinguished from allied
genera by its sessile spikelets, a turbinate callus with a central
knob, an ovate or oblong lower glume that is crustaceous,
smooth or with flaps or columns, and that has broad
membranous wings at the apex. It also has a 1-2 awned
pedicelled spikelet that is as large as the sessile spikelet and a
pedicel fused to internodes (Fonseca, 2003). Fonseca (2003)
distinguished two groups within the genus based on the
morphological characters of the lower glume, leaf anatomical
characters and phytogeography.
The genus Ophiuros Gaertn. (subtribe Rottboellinae)
comprises four species distributed in Asia, Africa and Australia
(Yadav, 2010). Among these only Ophiuros bombaiensis is
endemic to Peninsular India, occurs in sympatry with
Glyphochloa (Kabeer & Nair, 2009) and also shows a lower
glume ornamentation similar to Glyphochloa (ornamentation
similar to Glyphochloa (Gosavi KVC, pers. comm.)). Therefore
we suspected it to be a species of Glyphochloa and wanted
to test its relationship with Glyphochloa. As a contrast, the
widespread O. exaltatus (L.) Kuntze was also included.
J
SE Journal of Systematics
and Evolution
XXX 2015 | Volume 9999 | Issue 9999 | 1–13 © 2015 Institute of Botany, Chinese Academy of Sciences
Spikelet, fruit/seed (caryopsis) and embryo morphology are
taxonomically significant in the higher-order classification
of grasses (Stebbins, 1956; Clayton & Renvoize, 1986). For
example, caryopsis morphology is important in the taxonomy
of Chloridoideae (Liu et al., 2005), tribe Hordeeae (Terrell &
Peterson, 1993) and Eleusine Gaertn. (Bin et al., 2011).
Although some morphological work has been performed on
Glyphochloa, so far no chromosome records exist for it.
Therefore, in this study we performed meiotic counts of
chromosome number and further considered the morphology
of its caryopsis and spikelet and its geographical distribution.
For the first time, we performed a phylogenetic analysis to
test the monophyly of the genus and its relationship to
O. bombaiensis, using nuclear (ITS; ribosomal internal
transcribed spacer) and plastid intergenic regions (atpB-
rbcL,trnT-trnL and trnL-trnF). The probable course of
evolution and diversification of Glyphochloa species and its
relationship with O. bombaiensis are presented in this paper,
based on these data.
Material and Methods
Specimen collection and vouchers
Plant material for the present investigation was collected
from mountainous regions of the Western Ghats and the
coastal regions of Konkan in India. The distribution data were
recorded as GPS co-ordinates. Live collected materials were
labeled and grown in the wire house of the Botanical Garden
in the Department of Botany, Shivaji University, Kolhapur,
Maharashtra, India. Spikelet morphology, caryopsis morphol-
ogy, meiotic chromosome counts and molecular work
were done from live material, consistently using the same
vouchered individuals. Herbarium voucher specimens are
deposited in the Department of Botany Herbarium (SUK),
Shivaji University, Kolhapur, Maharashtra, India (Table 1).
DNA extraction and amplification
DNA was extracted from about 20 mg dried leaf or spikelet
tissues by grinding in a mortar and pestle with acid-washed
sand and following the manufacturer’s protocol for the
Geneipure Plant genomic DNA extraction kit (Genei, Banga-
lore, India). The DNA was eluted in 100 mL of the elution buffer
provided. Each 25 mL PCR reaction consisted of 10buffer,
3 mmol/L MgCl
2
, 0.4 mmol/L dNTP mix, 10 nmol of each primer,
1 U Taq polymerase (Genei, Bangalore, India) and 1 mL of the
extracted DNA template (varying from 10 to 200 ng). Nuclear
ITS (ribosomal internal transcribed spacer) sequences were
amplified using primers 17SE and 26SE primers (Sun et al.,
1994). Plastid (ptDNA) trnT-trnL and trnL-trnF (partial trnL,
trnL gene, intergenic spacer between trnL and trnF, and
partial trnF) regions were amplified using primer pairs a/b and
c/f respectively (Taberlet et al., 1991). The plastid atpB-rbcL
spacer region was amplified using the primer pair atpB-1/rbcL-1
(Chiang et al., 1998). Thermocycler condition varied for each
gene region. For ITS, a touchdown PCR (polymerase chain
reaction) was performed with the following conditions: initial
denaturation of 95 °C for 2 min, followed by 15 cycles of 95 °C
for 30 s, 61 °C for 30 s with a decrease in 0.5 °C per cycle, 72 °C
for 1 min, followed by 20 cycles of 95°C for 20 s, 57°C for 30 s,
72 °C for 1 min and a final extension at 72 °C for 5 min. For the
trnT-trnL region, thermocycler conditions were as follows:
initial denaturation at 95 °C for 2 min, followed by 35 cycles of
95 °C for 30 s, 50°C for 30 s, extension at 72 °C for 1min and
afinal extension at 72 °C for 5 min. For the trnL-trnF region,
thermocycler conditions were as above except that the
annealing temperature was 58 °C. For atpB-rbcL an annealing
temperature of 52 °C was used in the same PCR program as
above.
Amplification products were checked on 1% agarose gel
stained with ethidium bromide and purified with Qiaquick PCR
purification kit (Qiagen, Hilden, Germany). The ITS regions of
some specimens showed multiple bands; bands of expected
size (with respect to unambiguous bands in other species)
were gel-purified before sequencing. Products were quanti-
fied approximately using the gel image with reference to the
DNA markers loaded and sent for sequencing to Eurofins
Genomics (Eurofins Genomics India Pvt. Ltd, Bangalore) using
the forward and reverse primers used for amplification. Failed
reactions were repeated with slightly modified PCR con-
ditions, fresh DNA extractions, or alternate PCR product
purification methods. The list of taxa used in this study and
their GenBank accession numbers is provided in Table 1.
All ITS, trnL-trnF and atpB-rbcL sequences available on
GenBank for tribe Andropogoneae were downloaded to build
preliminary phylogenetic trees (one analysis per region). We
refer to these as the large taxon sets. We used these trees to
sub-sample taxa with close relationships to Glyphochloa for
more focused phylogenetic analyses (Table S1). For one
accession of specimen Ophiuros exaltatus (L.) Kuntze (GVC
Gosavi 2888), it was only possible to sequence the ITS region
by cloning the PCR product and subsequently sequencing the
plasmid insert using standard primers. Four clones were
sequenced, and as they showed minor variation to each other
they were all included in the large ITS dataset. Cloning and
sequencing were done commercially by Amnion Biosciences
(Amnion Biosciences Pvt. Ltd. Bangalore). Amplification of the
plastid markers was mostly successful but sequencing of
some samples failed because of long stretches of AT regions in
the intergenic regions and these regions could not be included
in the analysis (see Table 1).
Sequence analysis and phylogeny
Sequence chromatograms were checked for quality and
forward and reverse sequences were assembled into contigs
using the pregap4 and gap4 modules of the Staden package v.
2.0.0b10 (Bonfield et al., 1995). Completed sequences were
aligned using the Clustal W algorithm (Thompson et al., 1994)
as implemented in BioEdit v. 7.1.3 (Hall, 1999). Sequences were
further checked manually and edited using Se-Al v. 2.0a11
(Rambaut, 2002). Gaps and ambiguously aligned regions were
eliminated before phylogenetic analysis. Sequence character-
istics such as variable sites and the number of parsimony
informative sites were assessed using PAUPv. 4 beta 10
(Swofford, 2003).
The best-fit model of nucleotide evolution for each of the
individual datasets and the plastid combined dataset (Table 2)
was determined with ModelTest v. 0.1.1 (Posada, 2008) using
the Akaike information criterion (AIC). A Bayesian tree and
associated branch support measures (posterior probabilities)
were obtained in MrBayes v. 3.1.2 (Ronquist & Huelsenbeck,
2003). Two simultaneous independent runs with four Markov
2 Gosavi et al.
J. Syst. Evol. 9999 (9999): 1–13, 2015 www.jse.ac.cn
Table 1 Voucher specimens and GenBank accession numbers of the gene regions used in this study. Spikelet morphology,
caryopsis morphology, meiotic counts and molecular work have been done from the same vouchers
No. Botanical name Voucher information ITS atpB-rbcL trnL-trnF trnT-trnL
1Glyphochloa acuminata
(Hack.) Clayton var.
acuminata
India, Maharashtra, Sindhudurg district,
Sawantwadi 16.11.2008 K.V.C. Gosavi
2935 (SUK)
KT387609 KT387658 KT387627 KT387643
2G. acuminata (Hack.)
Clayton var. stocksii
(Hook.f.) Clayton
(Hook.f.) W.D.Clayton India, Maharashtra,
Ratnagiri district, Pali 28.10.2007 K.V.C.
Gosavi 2826 (SUK)
KT387610 KT387659 KT387628 KT387644
3G. divergens (Hack.) Clayton India, Maharashtra, Sangli district, Chandoli
11.11.2006 K.V.C. Gosavi 2748 (SUK)
KT387611 - KT387629 KT387645
4G. forficulata (C.E.C.Fischer)
Clayton
India, Maharashtra, Kolhapur district,
Radhanagari Shelap 18.10.2007 K.V.C.
Gosavi 2806 (SUK)
KT387612 KT387660 KT387630 KT387646
5G. forficulata (C.E.C.Fischer)
Clayton
India, Maharashtra, Sindhudurg district,
Chaukul 4.10.2008 K.V.C. Gosavi 2925
(SUK)
--- -
6G. forficulata (C.E.C.Fischer)
Clayton
India, Maharashtra, Kolhapur district,
Malakapur, Pandhapani 26.09.2008 K.V.C.
Gosavi 2914 (SUK)
--- -
7G. goaensis (Rolla Rao &
Hemadri) Clayton
India, Goa Tisk-Usgaon 16.09.2008 K.V.C.
Gosavi 2905 (SUK)
KT387613 KT387661 KT387631 KT387647
8G. henryi Janarth., Joshi &
Rajkumar
India, Goa Tisk-Usgaon 18.09.2008 K.V.C.
Gosavi 2906 (SUK)
KT387614 - KT387632 -
9G. maharashtraensis Potdar
& S.R. Yadav
Kolhapur district, Gargoti, Kondushi
plateau 17.10.2007 K.V.C. Gosavi 2797
(SUK)
KT387620 KT387663 KT387638 KT387652
10 G. mysorensis (Jain &
Hemadri) Clayton
India, Maharashtra, Kolhapur district,
Borbet 07.10.2007 K.V.C. Gosavi 2787
(SUK)
KT387611 KT387662 KT387633 KT387648
11 G. ratnagirica (Kulkarni &
Hemadri) Clayton
India, Maharashtra, Sindhudurg district,
Chaukul 27.11.2006 K.V.C. Gosavi 2742
(SUK)
KT387617 - KT387634 -
12 G. ratnagirica (Kulkarni &
Hemadri) Clayton
India, Karnataka, Belgam district, Amgaon
plateau 30.10.2007 A. N. Chandore
ANC-931 (SUK)
--- -
13 G. santapaui (Kulkarni &
Deshpande) Clayton
India, Maharashtra, Ratnagiri district,
Ratnagiri 14.09.2008 K.V.C. Gosavi 2903
(SUK)
KT387617 - KT387635 KT387649
14 G. talbotii (Hook.f.) Clayton India, Goa, Goa University 16.09.2008 K.V.C.
Gosavi 2904 (SUK)
KT387618 - KT387636 KT387650
15 G. veldkampii M. A. Fonseca
& Janarthanam
India, Goa, Mollem 28.11.2008 K.V.C. Gosavi
2944 (SUK)
KT387619 - KT387637 KT387651
16 Ophiuros bombaiensis Bor India, Maharashtra, Sindhudurg district,
Sawantwadi 28.10.2006 K.V.C. Gosavi
2699 (SUK)
KT387621 KT387664 KT387639 KT387653
17 O. exaltatus (L.) Kuntze India, Maharashtra, Kolhapur district,
Kagal 15.08.2008 K.V.C. Gosavi 2888
(SUK)
KT387622 KT387665 KT387640 KT3876454
18 O. exaltatus (L.) Kuntze India, Madhya Pradesh, Burhanpur district,
Nepanagar 23.2.2013 K.V.C. Gosavi 002
(SUK)
KT387623 - - -
19 Hackelochloa granularis (L.)
Kuntze
India, Nadurbar district, Shahada
16.10.2014 K.V.C Gosavi 639 (SUK)
KT387624 KT387666 KT387641 KT387655
20 Manisuris myurus L. India, Tamilnadu, Sattur 3.11.2014 K.V.C.
Gosavi 642 (SUK)
KT387626 KT387668 - KT387657
21 Mnesithea veldkampii
Potdar, S.P.Gaikwad,
Salunkhe & S.R.Yadav
India, Maharashtra, Kolhapur district,
Panhala, Masai plateau 31.12.2006 K.V.C.
Gosavi 2757 (SUK)
KT387625 KT387667 KT387642 KT387656
-, missing data. SUK, Shivaji University Kolhapur.
Molecular phylogeny of Glyphochloa 3
www.jse.ac.cn J. Syst. Evol. 9999 (9999): 1–13, 2015
chains were run for 2 million generations, and trees were
sampled every 100 generations, resulting in 20 000 trees. The
convergence of the runs and were checked using Tracer v 1.6.0
(Rambaut et al., 2014). The effective sample size (ESS) values
for each parameter of the run were greater than 200 after 2
million runs. The first 4000 generations were discarded as
burn-in and a majority-rule consensus tree was obtained based
on the remaining 16 000 trees. Maximum likelihood trees and
support values (bootstrap support; BS) were obtained using
RAxML (Stamatakis, 2006) using raxmlGUI v. 1.3 (Silvestro &
Michalak, 2012). One thousand bootstrap replicates were
performed using the standard bootstrap method in RAxML.
Trees were visualized using Figtree v. 1.4.0 (Rambaut, 2012).
Bayesian posterior probability (PP) 0.99 and ML bootstrap
support (BS) 90% were considered to be well supported
branches; PP 0.95–0.98 and BS 70–89% values were consid-
ered to indicate moderate support, and PP <0.95 and BS <70%
values were considered to indicate poorly supported
branches.
We ran two distinct versions of the phylogenetic analyses.
For the first set (which we refer to as the ‘large taxon sets’)
we considered a broad taxon sampling across the tribe
Andropogoneae for each of three regions (the nuclear ITS
region, and two plastid regions, trnL-trnF and atpB-rbcL)in
order to assess: (i) the monophyly of Glyphochloa,and(ii)to
infer which taxa are most closely related to Glyphochloa.For
the second set (which we refer to as the ‘small taxon sets’),
we performed more taxonomically focused phylogenetic
analyses, considering taxa closely related to Glyphochloa
as identified from the large samples. New alignments
were prepared for these, so we were also interested to
see if this difference had an effect on local phylogenetic
relationships.
Spikelet morphology
Spikelets were studied under a stereo-zoom microscope with
critical examination of the lower glume of sessile spikelet.
Caryopsis and embryo morphology
The length of the caryopses and the embryo within were
studied by soaking the seeds in distilled water for at least 12 h;
observations and photographs were taken with a Nikon SMZ
stereo-microscope. Embryos were visible without dissection
of the fruit coat.
Cytology
Squash preparations were made from young spikelets fixed in
Conroy’sfixative (ethanol and glacial acetic acid, 3:1). Anthers
were smeared in propionic-orcein and analyzed for meiotic
chromosome counts. Photographs were taken with a Carl-
Zeiss Jenaval compound microscope.
Results
Molecular phylogeny
Large taxon sets
The ‘large taxon’ITS dataset consisted of 196 taxa (including
13 ingroup taxa from this study). This dataset consisted of 708
aligned characters, of which 275 variable characters were
parsimony-informative (Table 2). The Bayesian phylogenetic
tree depicted Glyphochloa as monophyletic if Ophiuros
bombaiensis is considered to be part of the genus (see
below), but without strong support (Fig. S1) However there
were two subclades that had strong Bayesian and bootstrap
branch support (Fig. S1), referred to as ‘group I’and ‘group II’
below.
The large trnL-trnF data set consisted of 113 taxa (Table 2;
including 13 ingroup taxa) based largely on GenBank
sequences of Andropogoneae. The large trnL-trnF dataset
consisted of 1040 aligned sequences of which 95 variable
characters were parsimony informative (Table 2). The trnL-
trnF region showed a poorly resolved tree (Fig. S2) which did
not retrieve Glyphochloa as monophyletic. However, several
Glyphochloa subclades formed a polytomy with Hackelochloa,
Mnesithea and O. exaltatus. As with the ITS dataset, the trnL-
trnF analysis also showed two well supported subclades of
Glyphochloa, one of which corresponded to group I (including
O. bombaiensis), and the other to the group II subclade
(except for G. ratnagirica, which did not group with either
clade).
As the trnT-trnL region was not available for most grasses in
GenBank, a large dataset of trnT-trnL could not be constructed
for it. The large atpB-rbcL dataset consisted of 106 taxa with
six ingroup taxa (for the atpB-rbcL region only six species of
Glyphochloa could be sequenced; Table 1). The dataset had 814
aligned nucleotide characters of which 145 were parsimony
informative. In the Bayesian tree Glyphochloa formed a major
polytomy with different species of Hemarthria and Hackelo-
chloa (Fig. S3). Within that polytomy Glyphochloa was
recovered as monophyletic with moderate to strong support
(BPP 1, BS 88) based on the six sampled species. Even with the
limited samples of Glyphochloa in this dataset, group I and II
subclades were each retrieved as well supported clades.
For all three regions analyzed for large data sets, Ophiuros
bombaiensis was nested within Glyphochloa. In the large taxon
Table 2 Sequence lengths, excluded characters, parsimony informative character and model of nucleotide evolution of the
datasets
Gene region No. of
taxa
Length Parsimony informative
characters
Total variable
characters
Model of nucleotide
evolution
ITS large dataset 196 708 275 371 GTR þIþG
ITS smaller dataset 37 623 161 241 GTR þIþG
trnL-trnF large dataset 113 1040 95 218 GTR þG
atpB-rbcL large dataset 106 814 145 276 GTR þG
plastid DNA combined (trnL-trnF,trnT-
trnL and atpB-rbcL)
32 2190 69 216 GTR þG
4 Gosavi et al.
J. Syst. Evol. 9999 (9999): 1–13, 2015 www.jse.ac.cn
data sets involving single plastid regions, the resolution of the
clades was poor, likely due to low amount of sequence
variation in these regions (Figs. S2, S3).
Smaller taxon data sets
ITS data
This subsampled dataset consisted of 37 taxa and 623 aligned
characters of which 161 variable characters were parsimony
informative (Table 2). The Bayesian ITS tree recovered
Glyphochloa as monophyletic but without strong branch
support (Fig. 1). Two groups represented by different
chromosome numbers (n¼6 for group I; n¼7 for group II)
were each retrieved as well supported clades (Fig. 1; Table 3).
Ophiuros bombaiensis was found nested in group I with
G. talbotii (Hook.f.) Clayton.
Group I consists of low elevation species with single-awned
lower glumes. Group I was further resolved into two well
supported subclades: one consisting of G. henryi Janarth., V.C.
Joshi & S. Rajkumar, G. veldkampii M.A. Fonseca & Janarth.,
G.talbotii and O. bombaiensis and the second consisting of
two varieties of G. acuminata (Hack.) Clayton (G. acuminata
var. acuminata and G. acuminata var. stocksii), G. goaensis (Rao
& Hemadri) Clayton and G. santapaui (Jain & Desh.) Clayton.
Group II consists of G. divergens (Hack.) Clayton, G. forficulata
(C.E.C. Fisch.) Clayton, G. maharashtraensis Potdar and S.R.
Yadav, G. mysorensis (Kulkarni & Hemadri) Clayton and
G. ratnagirica (B. G. Kulk. & Hemadri) Clayton, which all
mostly occur in higher elevation lateritic plateaus, and which
have two-awned lower glumes (Table 3).
Plastid data
We performed additional analyses using a smaller number of
taxa, with all three plastid intergenic regions (atpB-rbcL,trnL-
trnF and trnT-trnL) concatenated, as they are expected to
Fig. 1. Bayesian phylogenetic inference showing relationships of Glyphochloa and relatives based on nuclear ITS data. Branches
with Bayesian posterior probabilities (PP) >0.95 are thickened. Support values are indicated beside branches (PP/ML bootstrap
support). Numbers in parentheses beside taxa indicate haploid chromosome number (x). Asterisk indicates polyploid species
(refer to Table 3 for actual chromosome number). Sequences generated for this study are indicated in bold red. MH and MP
indicate different accessions of Hackelochloa granularis. Scale ¼number of substitutions per site.
Molecular phylogeny of Glyphochloa 5
www.jse.ac.cn J. Syst. Evol. 9999 (9999): 1–13, 2015
Table 3 Summary of characters for Group I and Group II species of Glyphochloa
Name Chromosome
number
Awn
number
Embryo length
coverage on seed
Occurrence altitude Group
Glyphochloa acuminata 7 Usually 1 Not full Low (coastal plains) I
G. henryi 6 1 Full Low (coastal plains) I
G. goaensis 6 1 Full Low (coastal plains) I
G. santapaui 6 1 Full Low (coastal plains) I
G. talbotii 6 1 Full Low (coastal plains) I
G. veldkampii 6 1 Full Low (coastal plains) I
G. divergens 7 2 Not full High (mountains) II
G. maharastraensis 7 2 Not full High (mountains) II
G. forficulata 7, 14, 21, 28 2 Not full Low (coastal plains)/High (mountains) II
G. mysorensis 7 2 Not full High (mountains) II
G. ratnagirica 7, 14 2 Not full Low (coastal plains)/High (mountains) II
Ophiuros bombaiensis 7 0 Not full Low (coastal plains) I
O. exaltatus 7 0 Not observed Low (coastal plains) Outgroup
Fig. 2. Bayesian phylogenetic inference showing relationships of Glyphochloa and relatives based on a concatenated plastid DNA
dataset (atpB-rbcL,trnT-trnL and trnL-trnF regions). Branches with Bayesian posterior probabilities (PP) >0.95 are thickened.
Support values are indicated beside branches (PP/ML bootstrap support). Numbers in parentheses beside taxa indicate haploid
chromosome number (x). Asterisk indicates polyploid species (refer Table 3 for actual chromosome number). Sequences
generated for this study are indicated in bold red. Scale ¼number of substitutions per site.
6 Gosavi et al.
J. Syst. Evol. 9999 (9999): 1–13, 2015 www.jse.ac.cn
have a shared species history. For most of the taxa the trnT-
trnL sequences were not available from GenBank, which we
treated as missing data in the analysis. This smaller plastid
dataset consisted of 32 taxa and 2190 aligned characters,
after deletion of several ambiguously aligned microsatellite
regions. This dataset had 69 variable characters which where
parsimony-informative (Table 2). A Bayesian phylogenetic tree
inferred from the combined ptDNA regions showed the group
I and group II subclades (the latter excluding G. ratnagirica)as
sister taxa, although their sister relationship was not well
supported (Fig. 2). Within group I, we recovered two well
supported clades, one consisting of O. bombaiensis,G. talbotii,
G. henryi,G. veldkampii and G. goaensis, and the second
consisting of G. santapaui and the two varieties of
G. acuminata. The positions of G. goaensis and G. ratnagirica
varied substantially between the plastid and nuclear data sets
(Figs. 1, 2). However, the placement of G. ratnagirica as the
sister group of group I and II taxa was not well supported by
the plastid data. Within group II, G. divergens,G. forficulata,
G. mysorensis and G. maharashtraensis formed a well
supported clade (Fig. 2). The main characteristics of group I
and group II taxa are summarized in Table 3.
Spikelet morphology
The genus Glyphochloa can be divided into two groups on
the basis of spikelet morphology, particularly the number of
awns on the lower glume of the sessile spikelet. One has
two awns (group II) and the other has a single awn (group I;
O. bombaiensis and sometimes G. acuminata have none).
Each species can be identified by the ornamentation pattern
on the lower glume of sessile spikelet (Fig. 3; Table 4).
Spikelet size, shape and the ornamentation pattern of the
lower glume of G.forficulata and G. mysorensis commonly
showed variation, even within a population. The lower
glume of sessile spikelets of G. forficulata and G. mysorensis
are variable in possessing few to many dense tubercles,
having different sizes of tubercles, no tubercles or
possessing only furrows in the place of tubercles. In
G. goaensis,G. henryi,G. santapaui,G. talbotii,andG.
veldkampii ornamentation was sometimes absent on the
lower glume of the sessile spikelet.
Glyphochloa acuminata has three varieties, G. acuminata var.
acuminata,G. acuminata var. stocksii and G. acuminata var.
woodrowii. These can be distinguished by the size of their
spikelets and the number of ridges present on the lower
glume. As with G. forficulata and G. mysorensis, there is
variation in the morphology of the lower glume of sessile
spikelets of G. acuminata. In addition, G. acuminata has 0 to 2
awns per spikelet. The absence of awn in the lower glume of
the sessile spikelets in G. acuminata is similar to that of
O. bombaiensis (Fig. 3).
Meiotic chromosome counts
In Glyphochloa we found two putative base chromosome
numbers for the clades recovered in the molecular phyloge-
netic analysis: x¼6 for most groupI taxa (G. goaensis,G. henryi,
G. santapaui,G. talbotii,andG. veldkampii)andx¼7 for group II
taxa (G. acuminata,G. divergens,G. forficulata,G. mysorensis,
G. ratnagirica and G. maharashtraensis) (Fig. S5). However,
despite being in group I in molecular analysis, G. acuminata and
O. bombaiensis had a chromosome number of 7 (Table 3; Figs. 3,
S4). In G. ratnagirica we found n¼7 and 14 in different
populations, and for G. forficulata we found different numbers
n¼7, 14, 21, 28 even within a single population.
Caryopsis and embryo morphology
In seed morphology, some variation in embryo length was
observed between group I and group II species (Table 3).
Group I taxa had an embryo extending for the full length of
the seed (G. goaensis,G. henryi,G. santapaui,G. talbotii, and
G. veldkampii). In O. bombaiensis (group I) the embryo length
was less than half the size of the caryopsis. The group II
species had an embryo length ranging from 50 to 80% of the
caryopsis length (Table 4).
Geographical distribution
The genus comprises 11 species and four varieties, all of which
occur in western Maharashtra and Goa states (Fig. 4) of India.
Fig. 3. Lower glume of Glyphochloa species. A,G. henryi.B,G. goaensis.C,G. veldkampii.D,G. talbotii.E,G. santapaui.F,Ophiuros
bombaiensis (Glyphochloa bombaiensis). G,G. acuminate.H,G. forficulata.I,G. mysorensis.J,G. ratnagirica.K,G. divergens.
L,G. maharashtraensis. Scale, one division ¼1 mm.
Molecular phylogeny of Glyphochloa 7
www.jse.ac.cn J. Syst. Evol. 9999 (9999): 1–13, 2015
Table 4 Lower glume, caryopsis, embryo characters and meiotic chromosome counts of Glyphochloa and Ophiuros bombaiensis
No. Taxon Lower glume of sessile Spikelet description Caryopses and embryo characters Meiotic
Counts (n)
1Ophiuros bombaiensis Bor 2.5–31–1.2 mm long, coriaceous, ovate–oblong, 7–9 nerved, glabrous, obscurely
pitted between the nerves, glabrous, keels narrowly winged on both sides at
the apex, awnless, apex acute.
1.8 0.8 mm, elliptic; embryo
0.8 mm long, less than half the
seed length.
7
2Glyphochloa acuminata (Hack.)
Clayton
4–11 1–1.6 mm, coriaceous, narrowly ovate, 7–9 nerved, with or without
transverse ridges, winged at the apex, 0–2 awned at apex.
1.2 0.35 mm, elliptic–oval; embryo
0.8 mm long, 2/3 the seed length.
7
3G. divergens (Hack.) Clayton 2.5–3.5 0.8–1 mm, coriaceous, ovate, 7–nerved, pitted on the back, winged in
the upper half, cleft at apex into two divaricate awns.
1.1 0.3 mm, oval–elliptic; embryo
0.55 mm long, 1/2 the seed length.
7
4G. forficulata (C.E.C.Fischer)
Clayton
6.5–11 1.2–3 mm, coriaceous, oblong–ovate, 5–6 nerved, with or without
tubercles, tubercles few to many, broadly winged on both sides at the apex, 2–
awned, awn 2–10 mm long.
1.4 0.45 mm, elliptic–oval; embryo
1.05 mm long, 5/7 the seed length.
7, 14, 21 & 28
5G. goaensis (Rolla Rao & Hemadri)
Clayton
8–15 1.8–2 mm, coriaceous, ovate–elliptic, glabrous or puberulous, 1–2 ridged
near base, ridges overlapping downwards, sometimes ridged absent, unequally
winged on the margins, 1–awned at apex, awn 5–10 mm long.
1.0 0.9 mm, rounded; embryo as
long as the seed.
6
6G. henryi Janarth., Joshi &
Rajkumar
8–15 1.8–2.5 mm, coriaceous, ovate–elliptic, 1–2 transverse ridge at near base
extending end to end, facing upward along the margin forming collar, rarely
transverse ridge absent, margins wavy, lateral edges ciliate, glabrous or
puberulous, winged on the margins, 1-awned at apex, awn 5–10 mm long.
1.2 1.1 mm, rounded; embryo as
long as the seed.
6
7Glyphochloa maharashtraensis
Potdar & S.R. Yadav
6–10 1.3–1.6 mm, coriaceous, elliptic, 4–5 continuous ridges pointing and
overlapping upwards, margins inflexed, 7–9 nerved, keels broadly winged on
both sides along the awns, 2-awned, awns 4–6 mm long.
1.0 0.7 mm, oval–elliptic; embryo
0.5 mm long, 1/2 the seed length.
7
8G. mysorensis (Jain & Hemadri)
Clayton
2–47–12 mm, coriaceous, ovate–elliptic, obscurely 7–nerved, tubercle many to
few, sometimes marginal tubercles longer than those at middle region or
middle tubercled absent marginal tubercled tuft of small cilia at their tips or
glabrous, cleft at the apex and bearing two short thinly membranous wings
and prolonged into 2 divergent 5–8 mm long awns.
1.4 0.45 mm, elliptic; embryo
0.84 mm long, 3/5 seed length.
7
9G. ratnagirica (Kulkarni &
Hemadri) Clayton
5–81–3 mm, ovate-oblong, 5–nerved, with or without few to many tubercles, 2-
keeled, keels winged, wing glabrous to hairy, cleft at apex, 2–awned, awns 4–
5 mm long.
1.2 0.4 mm, elliptic–oval; embryo
0.7 mm long, 3/5 the seed length.
7&14
10 G. santapaui (Kulkarni &
Deshpande) Clayton
5–81–3 mm, ovate-oblong, 5–nerved, with or without few to many tubercles, 2–
keeled, keels winged, wing glabrous to hairy, cleft at apex, 2–awned, awns 4–
5 mm long.
0.45 0.45 mm, rounded; embryo as
long as the seed.
6
11 G. talbotii (Hook.f.) W. D. Clayton 10–12 0.8–1 mm, coriaceous, elliptic, with 2–3 transverse ridges at the base,
ridges pointing and overlapping upwards, 2–keeled, winged on the keels, apex
prolonged with 6–8 mm long slender single awn.
1.4 1.0 mm, oval-rounded; embryo
as long as the seed.
6
12 G. veldkampii M. A. Fonseca &
Janarthanam
10–15 0.8–1 mm, coriaceous, elliptic, with 9–10 transverse ridges at the base,
ridges pointing and overlapping upwards, cup like, erect–patent with gap
between the glume surface and collar, broken at one side, hairy at lateral edge,
rarely with two laterally hooks, winged on the keels, apex prolonged with 6–
8 mm long slender single awn.
1.2 0.8 mm, oval-rounded; embryo
as long as the seed.
6
8 Gosavi et al.
J. Syst. Evol. 9999 (9999): 1–13, 2015 www.jse.ac.cn
Of those, G. santapaui,G. ratnagirica,G. maharashtraensis and
G. maharashtraensis var. hirsuta are endemic to Maharashtra
state, while G. henryi and G. veldkampii are endemic to Goa.
Glyphochloa goaensis and G. talbotii are distributed in
Maharashtra and Goa states only (Fig. 5). Glyphochloa
forficulata is comparatively widely distributed and also occurs
in Madhya Pradesh (Fonseca, 2003), Karnataka and Kerala
states (Shreekumar & Nair, 1991). Glyphochloa divergens and
G. acuminata are also distributed in Karnataka and Kerala
states (G. divergens is represented by a single collection in
Kerala state; Sreekumar & Nair, 1991). Glyphochloa mysorensis
and G. ratnagirica are also distributed in Karnataka (Malpure
NV, unpublished data).
Species of Glyphochloa also exhibit an altitudinal gradation
in their distribution. Glyphochloa acuminata,G. goaensis,
G. henryi,G. santapaui and G. talbotii are confined to low
altitude (0–300 m) lateritic plateaus. Glyphochloa divergens,
G. mysorensis, and G. maharashtraensis are distributed in
higher altitude lateritic plateaus (600–1200 m). Glyphochloa
ratnagirica and G. forficulata are found from the coastal plains
to hills (0–1200 m) in Northern Western Ghats (Sahyadris).
Ophiuros bombaiensis is distributed in Maharashtra, Goa (Joshi
& Janarthanam, 2004), Karnataka and Tamil Nadu (Nayar &
Ramamurthy, 1976) in low altitudes.
Discussion
In phylogenetic trees based on the molecular data, species of
Glyphochloa generally were recovered together as a mono-
phyletic group, that comprises two subclades (Figs. 1, 2, S1–
S3), if Ophiuros bombaiensis is considered as part of the genus
(see below). However, G. ratnagirica is weakly supported as
the sister group of the other species in the tree inferred from
the concatenated plastid DNA data (Fig. 2), The sister-group
relationship of group I and II species (with or without
G. ratnagirica) was poorly supported (e.g., Figs. 1, 2), but
characteristics such as the number of awns on the lower
glume of the sessile spikelet, embryo size and basic
chromosome number are generally consistent with these
two groups (Fig. 3; Tables 3, 4).
Ophiuros bombaiensis is nested within Glyphochloa in all of
our phylogenetic analyses. In contrast, another species
O. exaltatus is distantly related to Glyphochloa according to
our nuclear and plastid phylogenies (e.g., Figs. 1, 2, S1–S3).
Ophiuros bombaiensis shares the following characters with
Glyphochloa,reflecting its placement in molecular analysis: (i)
an annual habit; (ii) restriction to lateritic plateaus of lower
altitude; (iii) a succulent habit similar to group I of Glyphochloa
species; (iv) the lower glume of the sessile spikelet is similar to
that of a G. acuminata variant lacking awns; (v) its caryopsis
and embryo characters are similar to group II species and (vi)
its chromosome number x¼7 is comparable to other group II
species that have double awns. We therefore consider it to be
a member of the genus (see below).
Glyphochloa group I species (O. bombaiensis,G. acuminata,
G. goaensis,G. henryi,G. santapaui,G. talbotii and G.
veldkampii) are restricted and adapted to plateaus of lower
altitude. The lower glume of the sessile spikelet is single-
awned in all the species except G. acuminata, in which it is
usually single-awned, although cases with two awns and no
awns were also observed. Except for G. acuminata and
Fig. 4. Geographical distribution of widely distributed Glyphochloa species.
Molecular phylogeny of Glyphochloa 9
www.jse.ac.cn J. Syst. Evol. 9999 (9999): 1–13, 2015
O. bombaiensis, all species of group I have a base chromosome
number of x¼6.
Glyphochloa group II (G. divergens,G. forficulata,
G. mysorensis,G. ratnagirica and G. maharashtraensis) species
are confined to higher altitude lateritic plateaus apart from
G. forficulata and G. ratnagirica, which also occur in low
elevation coastal plateaus. All species of group II have a
putative base chromosome number of x¼7. Glyphochloa
mysorensis showed different morphotypes but a single
cytotype. The widespread species G. forficulata showed
difference in morphotypes and in ploidy level (Table 4). The
group I and II subclades occur in allopatry, further supporting
their distinctiveness.
The entire genus groups into a poorly supported clade
within tribe Andropogoneae phylogeny in both ITS and plastid
analysis (Figs. 1, 2, S1, S2). The lack of support for the
monophyly of Glyphochloa might be due to the scarcity of
phylogenetically informative variable characters in the
alignment. Based on the plastid phylogeny, G. ratnagirica
may represent a lineage distinct from group I and II taxa (cf.,
Figs. 2, S2, S3). The addition of additional sequences might
increase the support for Glyphochloa as a clade, but the weight
of evidence at present does not warrant recognition of the
two clades as distinct genera.
In both the ITS and plastid phylogeny members of
Hackelochloa granularis,Hemarthria,Mnesithea,Manisuris
were found to be closely related to Glyphochloa (Figs. 1, 2),
possibly along with other genera (Fig. 2), including Ophiuros s.-
s. (represented by O. exaltatus here). These relationships are
not well-supported by either data sets, however, and the
precise relationships varied between the nuclear and plastid
trees. Hackelochloa granularis is an annual grass spread across
the tropics (GBIF, 2014) and it shares some similarities with
Glyphochloa, such as having paired spikelets, one of which is
sessile and the other pedicelled (also found in Rottboellia L.f.
and Mnesithea Kunth.). However, in contrast to Glyphochloa,
the lower glume of sessile spikelet in Hackelochloa granularis is
not awned (winged) and is orbicular in shape. As with
G. ratnagirica, the lower glume of Hackelochloa granularis is
heavily ornamented (Yadav, 2010). Hemarthria,Mnesithea and
Manisuris are old world grasses occurring mostly in Southeast
Asia, suggesting a close biogeographic affinity. The current
data shows that Hackelochloa,Mnesithea,Hemarthria and
Manisuris are probably the closest relatives of Glyphochloa,
although several other genera cannot be excluded (Kerrio-
chloa,Rottboellia,Coelorachis; Fig. 1).
Hybridization across species and genera is common in
grasses (Cayouette & Darbyshire, 1993; Levy & Feldman,
2002) and experimental intergeneric hybrids are also
reported to produce viable offspring (Griffin et al., 2011).
As a result, highly divergent ITS polymorphism is a not
uncommon phenomenon in grasses (Hershkovitz et al.,
1999; Alvarez & Wendel, 2003). Incomplete homogenization
of nuclear ribosomal DNA regions including ITS suggests
recent allopolyploidy and is useful in detecting hybridization
and for understanding genome evolution in polyploid
species (Ainouche & Bayer, 1997; Hershkovitz et al., 1999;
Yonemori et al., 2002; Zhang et al., 2002; Liu et al., 2006). As
polymorphic forms of ITS sequences were observed in
O. exaltatus, it might be a recently evolved allopolyploid
species; however, the isoforms of ITS region were very
closely related and did not occupy different clades in the
phylogeny.
The nuclear ribosomal ITS region has been widely used as a
barcode for plant species, although there are failures in some
cases, as shown by hard-to-read sequences that reflect
polymorphism within species (Hollingsworth, 2011; Li & China
Plant BOL Group, 2011). The ITS region, a part of the nuclear
rDNA operon, is represented by numerous copies in the cell
and therefore duplications in some regions can lead to
pseudogenes that may not completely decay. In addition to
lateral gene transfer, reflecting hybridization and introgres-
sion events, the plastid and nuclear markers may also show
difference in topologies because of different evolutionary
rates and incomplete lineage sorting (Degnan & Rosenberg,
2009). However, in our study, both the markers were fairly
Fig. 5. Geographical distribution of narrowly distributed
Glyphochloa species. Light grey indicates low elevation coastal
areas (0–300 m), dark grey indicates higher elevation plateaus
(600–1200 m).
10 Gosavi et al.
J. Syst. Evol. 9999 (9999): 1–13, 2015 www.jse.ac.cn
congruent, adding support to there being two major lineages
in Glyphochloa.
As group I species and several species of group II
(G. acuminata and G. bombaiensis) have a basic chromosome
number of x¼7, the ancestral chromosome number of
Glyphochloa may be x¼7. Hackelochloa granularis also has a
base chromosome number of n¼7 (Ashan et al., 1994). If so,
the base chromosome number of x¼6 in group I would be a
derived character reflecting nullisomic aneuploidy. Most
genera in Panicoideae have base chromosome numbers of
x>9 (Ashan et al., 1994). A basic chromosome number of
x¼6inGlyphochloa group I indicates a relatively compact
arrangement of the nuclear genome in this clade.
Glyphochloa and Ophiuros were not included in a recent
study of subfamily Panicoideae (Teerawatananon et al., 2011).
Within Andropogoneae, the members of the subtribe
Rottboellinae are polyphyletic, and a revision of taxa in this
subtribe is needed if a phylogenetic classification is to be
followed. A genus-level phylogeny of several closely related
taxa of subtribe Rottboellinae is needed to further study the
relationship between Ophiuros,Rottboellia,Hackelochloa and
Glyphochloa, perhaps using additional plastid markers such as
matK.
Conclusions
The evolutionary history of Glyphochloa indicates that it
comprises two major lineages (with the possible exception of
G. ratnagirica). One clade (group I: G. acuminata,G. goaensis,
G. henryi,G. santapaui,G. talbotii and G. veldkampii) is adapted
to lower altitude (up to 300 m) and is characterized by a
single awn on the lower glume of sessile spikelets, an embryo
occupying the full length of the caryopsis, and a base
chromosome number x¼6 (except G. acuminata and
Ophiuros bombaiensis). The second clade (group II:
G. divergens,G. forficulata,G. maharashtraensis G. mysorensis
and possibly G. ratnagirica) is adapted to higher altitude (600
to 1200 m) lateritic plateaus and is characterized as having two
awns on the lower glume of sessile spikelets, an embryo
occupying 1/3 length of caryopsis, and a base chromosome
number x¼7. Ophiuros bombaiensis is nested in the group I
subclade in our molecular phylogenies, a result that is also
supported by spikelet morphology, caryopsis morphology,
base chromosome number and geographical distribution.
Therefore we make a new nomenclatural combination for
O. bombaiensis.
Taxonomy
A major taxonomical implications of this study is that Ophiuros
bombaiensis should be merged under genus Glyphochloa
based on our molecular phylogenetic results and shared
morphological characters, such as low plant height (vs. tall
height of O. exaltatus) and annual habit (vs. perennial habit of
O. exaltatus).
Glyphochloa bombaiensis (Bor) Gosavi, S. R. Yadav, Praveen
Karanth & S. Siddharthan, comb. nov. Ophiuros bombaiensis
Bor, Kew Bull. 1951: 167. 1951. Holotype: India, Mysore,
Talguppa, 600–1000 m, 1908-10, A. Meebold 10764 (K!).
Key to the species of Glyphochloa
1a. Lower glume of sessile spikelet without awns . . . . . . 2
1b. Lower glume of sessile spikelet with awns . . . . . . . . 3
2a. Pedicelled spikelet reduced . . . . . . . . . G. bombaiensis
2b. Pedicelled spikelet well developed . . . . . . G. acuminata
3a. Lower glume of sessile spikelet with single awn . . . . 4
3b. Lower glume of sessile spikelet with two divergent
awns .....................................9
4a. Lower glume of sessile spikelet flat on back, devoid of
tubercles or depressions, densely ciliate. . . . . . . . . . . .
................................. G. santapaui
4b. Lower glume of sessile spikelet concave or flat on
back, with tubercles or depressions, glabrous or
pubescent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
5a. Back of the lower glume with vertical and 3-4 transverse
ridges . . . . . . . . . . . . . . . . . . . . . . . . . . . . G. acuminata
5b. Back of the lower glume with only transverse ridges . . .
............................................ 6
6a. Transverse ridges on glume curved downwards . . . . . .
..................................G. goaensis
6b. Transverse ridges on glume curved upward . . . . . . . . 7
7a. Ridges on back of the lower glume of sessile spikelet
straight to form one to two collar like structures (Refer
Fig. S5) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . G. henryi
7b. Ridges on back of the lower glume of sessile spikelet
convex....................................8
8a. Single convex ridge on lower glume forming a flap like
structure . . . . . . . . . . . . . . . . . . . . . . . . G. veldkampii
8b. 1–3 convex ridges on lower glume without flap like but
overlapping . . . . . . . . . . . . . . . . . . . . . . . . . G. talbotii
9a. Lower glume of sessile spikelet with continuous 2–5
ridges, without any hooks . . . . . . G. maharashtraensis
9b. Lower glume of sessile spikelet without ridges and with
or without hooks . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
10a. Lower glume of sessile spikelet oblong or obovate,
hooks about 0.5mm long or less or without hooks . . . 11
10b. Lower glume of sessile spikelet orbicular, with about
0.75 mm long hooks . . . . . . . . . . . . . . . . G. mysorensis
11a. Lower glume up to 5 mm long . . . . . . . . . G. divergens
11b. Lower glume 7–16 mm long . . . . . . . . . . . . . . . . . . . .12
12a. Leaves papery; lower glume of sessile spikelet with
tubercles and/or ridges, rarely smooth. . . .G. forficulata
12b. Leaves succulent; lower glume of sessile spikelet
conspicuously pitted and tuberculate but without ridges,
sometimes smooth. . . . . . . . . . . . . . . . . G. ratnagirica
Acknowledgements
We thank our funding agencies Department of Biotechnology
(DBT), Council of Scientific and Industrial Research (CSIR)
(CSIR-SRA Pool No. 8492), University Grants Commission
(UGC), Science and Engineering Research Board (SERB) (Fast
track grant SERC/LS-296/2011 and SB/FT/LS-130/2012) and the
Ministry of Environment and Forests (MOEF) for financial
assistance. We are thankful to authorities of the University
and Head, Department of Botany for providing facilities, to
Prof. M. K. Janarthanam, Department of Botany, Goa
University, Goa for suggestions and encouragement. Authors
Molecular phylogeny of Glyphochloa 11
www.jse.ac.cn J. Syst. Evol. 9999 (9999): 1–13, 2015
also thank Mr. Makarand Aitawade, Department of Botany,
Shivaji University, Kolhapur and Mr. Anup Deshpande,
Department of Botany, Goa University, Goa for help. We
thank the two anonymous reviewers for valuable comments
and improving the manuscript. The authors declare no conflict
of interest.
References
Ainouche ML, Bayer RJ. 1997. On the origins of the tetraploid Bromus
species (section Bromus, Poaceae): Insights from the internal
transcribed spacer sequences of nuclear ribosomal DNA. Genome
40: 730–743.
Alvarez I, Wendel JF. 2003. Ribosomal ITS sequences and plant
phylogenetic inference. Molecular Phylogenetics and Evolution 29:
417–434.
Ashan SMN, Vahidi AA, Ali SI. 1994. Chromosome numbers and
incidence of polyploidy in Panicoideae (Poaceae) from Pakistan.
Annals of the Missouri Botanical Garden 81: 775–783.
Bin J, Peterson PM, Qing L. 2011. Caryopsis micromorphology of
Eleusine Gaertn. (Poaceae) and its systematic implications. Journal
of Tropical and Subtropical Botany 19: 195–204.
Blatter E, McCann C. 1935. The Bombay grasses. Scientific Monograph
No. 5. Imperial Council of Agricultural Research, New Delhi, India 21:
324.
Bonfield JK, Smith KF, Staden R. 1995. A new DNA sequence assembly
program. Nucleic Acids Research 23: 4992–4999.
Bor NL. 1960. The grasses of Burma, Ceylon, India, and Pakistan. Oxford:
Pergamon Press.
Cayouette J, Darbyshire SJ. 1993. The intergeneric hybrid grass “Poa
labradorica”.Nordic Journal of Botany 13: 615–629.
Chiang T-Y, Schall BA, Peng C-I. 1998. Universal primers for
amplification and sequencing a noncoding spacer between
atpB and rbcL genes of chloroplast DNA. Botanical Bulletin of
Academia Sinica 39: 245–250.
Clayton WD. 1980. Notes on the Tribe Andropogoneae (Gramineae).
Kew Bulletin 35: 813–818.
Clayton WD, Renvoize SA. 1986. GeneraGraminum: Grasses of the World.
(Kew Bulletin Additional Series XIII). Kew: Royal Botanic Garden.
Datta-Roy A, Karanth KP. 2009. The Out-of-India hypothesis: What do
molecules suggest? Journal of Biosciences 34: 687–697.
Degnan JH, Rosenberg NA. 2009. Gene tree discordance, phyloge-
netic inference and the multi-species coalescent. Trends in Ecology
and Evolution 24: 332–340.
Fonseca MA. 2003. Systematic studies on the genus Glyphochloa. Ph.D.
Thesis. Goa: University of Goa.
GBIF. 2014. Global biodiversity information facility. Distribution record
for Hackelochloa granularis (L.) Kuntze [online]. Available from
http://www.gbif.org/species/2705462 [accessed 12 August 2014].
Griffin PC, Robin C, Hoffmann AA. 2011. A next-generation sequencing
method for overcoming the multiple gene copy problem in
polyploidy phylogenetics, applied to Poa grasses. BMC Biology 9: 19.
Hall T. 1999. BioEdit: A user-friendly biological sequence alignment
editor and analysis program for Windows 95/98/NT. Nucleic Acids
Symposium Series 41: 95–98.
Hershkovitz MA, Zimmer EA, Hahn WJ. 1999. Ribosomal DNA
sequences and angiosperm systematics. In: Hollingsworth PM,
Bateman RM, Gornall RJ eds. Molecular systematics and plant
evolution. London: Taylor & Francis. 268–326.
Hollingsworth PM. 2011. Refining the DNA barcode for land plants.
Proceedings of the National Academy of Sciences USA 108:
19451–19452.
Hooker JD. 1894. Flora of British India VI. L. London: Reeve and
Company.
Irwin SJ, Narasimhan D. 2011. Endemic genera of angiosperms in India:
A review. Rheedea 21: 87–105.
Jain SK. 1972. The genus Manisuris L. (Poaceae) in India. Bulletin of the
Botanical Survey of India 12: 6–17.
Jain SK, Deshpande UR. 1969. Manisuris santapaui Jain et Deshpande,
A new grass from India. Bulletin of the Botanical Survey of India 10:
277–279.
Jain SK, Hemadri K. 1969. Manisuris mysorensis Jain et Hemadri, a new
grass from India. Bulletin of the Botanical Survey of India 10:
280–282.
Joshi VC, Janarthanam MK. 2004. The diversity of life-form type, habitat
preference and phenology of the endemics in the Goa region of
the Western Ghats, India. Journal of Biogeography 3: 1227–1237.
Kabeer KAA, Nair VJ. 2009. Flora of Tamil Nadu - Grasses. Botanical
Survey of India. Calcutta: Botanical Survey of India.
Karanth KP. 2015. An island called India: Phylogenetic patterns across
multiple taxonomic groups reveal endemic radiations. Current
Science 108: 1847–1851.
Kulkarni BG, Hemadri K. 1974. Manisuris ratnagirica Kulkarni et
Hemadri new grass from the Sahyadri range, Maharashtra state.
Indian Forester 100: 250–254.
Levy AA, Feldman M. 2002. The impact of polyploidy on grass genome
evolution. Plant Physiology 130: 1587–1593.
Li DZ, China Plant BOL Group. 2011. Comparative analysis of a large
dataset indicates that internal transcribed spacer (ITS) should be
incorporated into the core barcode for seed plants. Proceedings
of the National Academy of Sciences USA 108: 19641–19646.
Liu Q, Zhao N-X, Hao G, Hu X-Y, Liu Y-X. 2005. Caryopsis morphology of
the Chloridoideae (Gramineae) and its systematic implications.
Botanical Journal of the Linnean Society 148: 57–72.
Liu Q, Ge S, Zhang X, Zhu G, Lu B-R. 2006. Phylogenetic relationships in
Elymus (Poaceae: Triticeae) based on the nuclear ribosomal
internal transcribed spacer and chloroplast trnL-F sequences. New
Phytologist 170: 411–420.
Mittermeier RA, Gill PR, Hoffmann M, Pilgrim J, Brooks T, Mittermeier
CG, Lamoreux J, da Fonseca GAB. 2004. Hotspots Revisited: Earths
biologically richest and most endangered terrestrial ecoregions.
Mexico City: CEMEX.
Nayar MP, Ramamurthy K. 1976. Third list of species and genera of
Indian Phanerogams not included in J.D. Hooker’s Flora of British
India (excluding Bangladesh, Burma, Ceylon, Malayan Peninsula
and Pakistan). Bulletin of the Botanical Survey of India 15: 204–234.
Posada D. 2008. jModelTest: Phylogenetic model averaging. Molecular
Biology and Evolution 25: 1253–1256.
Rambaut A. 2002. SE-AL Sequence Alignment Editor, v2.0a11. Oxford:
University of Oxford.
Rambaut A. 2012. FigTree [online]. Available from http://tree.bio.ed.
ac.uk/software/figtree/ [accessed 7 August 2014].
Rambaut A, Suchard MA, Xie D, Drummond AJ. 2014. Tracer v1.6
[online]. Available from http://beast.bio.ed.ac.uk/Tracer [ac-
cessed 7 August 2014].
Rao RS, Hemadri K. 1968. A new species of Manisuris Linn. (Poaceae)
from Goa. Bulletin of the Botanical Survey of India 10: 106–109.
Ronquist F, Huelsenbeck JP. 2003. MrBayes 3: Bayesian phylogenetic
inference under mixed models. Bioinformatics 19: 1572–1574.
12 Gosavi et al.
J. Syst. Evol. 9999 (9999): 1–13, 2015 www.jse.ac.cn
Royal Botanic Gardens, Kew. 2013. Electronic Plant Information Center
[online]. Available from http://epic.kew.org/epic/ [accessed 20
March 2013].
Shreekumar PV, Nair VJ. 1991. Flora of Kerala Grasses. Calcutta:
Botanical Survey of India.
Silvestro D, Michalak I. 2012. raxmlGUI: A graphical front-end for
RAxML. Organisms Diversity and Evolution 12: 335–337.
Stamatakis A. 2006. RAxML-VI-HPC: Maximum Likelihood-based
phylogenetic analyses with thousands of taxa and mixed models.
Bioinformatics 22: 2688–2690.
Stebbins GL. 1956. Cytogenetics and evolution of the grass family.
American Journal of Botany 43: 890–905.
Sun Y, Skinner DZ, Liang GH, Hulbert SH. 1994. Phylogenetic analysis of
Sorghum and related taxa using internal transcribed spacers of
nuclear ribosomal DNA. Theoretical and Applied Genetics 89:
26–32.
Swofford DL. 2003. PAUP. Phylogenetic Analysis Using Parsimony
(and Other Methods), vers. 4.0b10. Sunderland: Sinauer
Associates.
Taberlet P, Gielly L, Pautou G, Bouvet J. 1991. Universal primers for
amplification of three noncoding regions of chloroplast DNA.
Plant Molecular Biology 17: 1105–1109.
Teerawatananon A, Jacobs SWL, Hodkinson TR. 2011. Phylogenetics of
Panicoideae (Poaceae) based on chloroplast and nuclear DNA
sequences. Telopea 13: 115–142.
Terrell EF, Peterson PM. 1993. Caryopsis morphology and classification
in the Triticeae (Pooideae: Poaceae). Smithsonian Contributions to
Botany 83: 1–25.
ThompsonJD,HigginsDG,GibsonTJ.1994.CLUSTALW:Improving
the sensitivity of progressive multiple sequence alignment
through sequence weighting, positions-specific gap penalties
and weight matrix choice. Nucleic Acids Research 22: 4673–
4680.
Welker CAD, Kellog EA, Prado J. 2014. Andropogoneae versus
Sacchareae (Poaceae: Panicoideae): The end of a great contro-
versy. Taxon 63: 643–646.
Yadav SR. 2010. Know your grass genera through hand lens. Kolhapur:
Shivaji University Press.
Yonemori K, Honsho C, Kanzaki S, Eiadthong W, Sugiura A. 2002.
Phylogenetic relationships of Mangifera species revealed by ITS
sequences of nuclear ribosomal DNA and a possibility of their
hybrid origin. Plant Systematics and Evolution 231: 59–75.
Zhang W, Qu LJ, Gu H, Gao W, Liu M, Chen J, Chen Z. 2002. Studies on
the origin and evolution of tetraploid wheats based on the
internal transcribed spacer (ITS) sequences of nuclear ribosomal
DNA. Theoretical and Applied Genetics 104: 1099–1106.
Supplementary Material
The following supplementary material is available online
for this article at http://onlinelibrary.wiley.com/doi/10.1111/
jse.12185/suppinfo:
Table S1. GenBank sequences used in focused phylogenetic
analyses with sequences of Glyphochloa generated here.
Fig. S1. Phylogram of a Bayesian tree obtained from nuclear
ITS sequences of Andropogoneae grasses (192 taxa). Numbers
in parentheses besides Glyphochloa species indicate haploid
chromosome number. Values on branches indicate Bayesian
probability support (PP) >0.5. Clades with >0.95 PP are
thickened. Sequences generated for this study are indicated in
red. Scale ¼number of substitutions per site.
Fig. S2. Phylogram of a Bayesian tree obtained from plastid
trnL-trnF region of Andropogoneae grasses (113 taxa).
Numbers in parentheses besides Glyphochloa species indicate
haploid chromosome number. Values on branches indicate
PP >0.5. Clades with >0.95 PP are thickened. Sequences
generated for this study are indicated in red. Scale ¼number
of substitutions per site.
Fig. S3. Phylogram of the Bayesian tree obtained atpB-rbcL
region of Andropogoneae grasses (106 taxa). Numbers in
parentheses besides Glyphochloa species indicate haploid
chromosome number. Values on branches indicate PP >0.5.
Clades with >0.95 PP are thickened. Sequences generated for
this study are indicated in red. Scale ¼number of substitutions
per site.
Fig. S4. Meiotic chromosomes: Bivalents of Glyphochloa
species and Ophiuros bombaiensis. Scale bar ¼11 mm. A,G.
henryi.B,G. goaensis.C,G.veldkampii.D,G. talbotii.E,G.
santapaui.F,O. bombaiensis (Glyphochloa bombaiensis). G,G.
acuminata.H,G. forficulata.I,G. mysorensis.J,G. ratnagirica.K,
G. divergens.L,G. maharashtraensis.
Fig. S5. Photograph of lower glume morphological characters
used in the key.
Molecular phylogeny of Glyphochloa 13
www.jse.ac.cn J. Syst. Evol. 9999 (9999): 1–13, 2015
