Content uploaded by Ross Barnett
Author content
All content in this area was uploaded by Ross Barnett on May 22, 2014
Content may be subject to copyright.
ORIGINAL PAPER
One Species or Two? Vicariance, Lineage Divergence and Low
mtDNA Diversity in Geographically Isolated Populations of South
Asian River Dolphin
G. T. Braulik &R. Barnett &V. Odon &V. Islas-Villanueva &
A. R. Hoelzel &J. A. Graves
#Springer Science+Business Media New York 2014
Abstract Despite their endangered status, the taxonomic re-
lationship between the two geographically isolated South
Asian river dolphin populations has never been comprehen-
sively assessed and remains contentious. Here we present the
first dedicated evaluation of the molecular phylogenetic rela-
tionship between the Indus (Platanista gangetica minor)and
Ganges (Platanista gangetica gangetica) River dolphins
using mitochondrial DNA from the control region and cyto-
chrome b, extracted from museum specimens. The 458 bp
partial control region sequences from 26 Indus River dolphin
samples exhibited no variation. Only six haplotypes were
identified in the 31 (18 Indus; 13 Ganges) complete (856 bp)
control region sequences obtained, none were shared between
subspecies, and there were five fixed differences between
them. Similarly low genetic diversity was found in a 541 bp
section of the cytochrome b gene (n=29). The lack of shared
haplotypes and fixed differences resulted in Φ
ST
for the partial
control region sequences of 0.932 (p<0.0001) and F
ST
of
0.843 (p<0.0001), indicating the long-term absence of gene
flow and clear genetic differentiation between the two geo-
graphically isolated populations. An externally calibrated mo-
lecular clock estimated that Indus and Ganges dolphins di-
verged around 550,000 years ago (95 % posterior probability
0.13–1.05 million years ago), possibly when dolphins from
the Ganges dispersed into the Indus during drainage capture.
Keywords Molecular phylogeny .Taxono my .Platanista
gangetica .Molecular diversity .Systematics .River capture
Introduction
The South Asian river dolphin is a relict species of cetacean,
the only extant member of one of the most basal cetacean
families, the Platanistidae. Despite the physical dissimilarity,
molecular analyses demonstrated that the Platanistidae family
is most closely related to the Kogiidae (dwarf and pygmy
sperm whales) and Physeteridae (sperm whales) (McGowen
et al. 2009; Steeman et al. 2009;Zhouetal.2011). Currently,
there are two closely related, but geographically isolated,
subspecies of South Asian river dolphin that are recognized;
the Indus River dolphin (Platanista gangetica minor)
inhabiting the Indus River system of Pakistan and India, and
the Ganges River dolphin (Platanista gangetica gangetica)in
the Ganges, Brahmaputra, and Karnaphuli River systems of
India, Bangladesh, and Nepal (Fig. 1). It was suggested that
these river dolphins evolved from a marine platanistid ances-
tor that colonized the shallow, epi-continental seas that pene-
trated low-lying areas of South Asia during a period of glob-
ally high sea levels in the Miocene, and as sea-levels receded
dolphins persisted in the freshwater river systems (Hamilton
et al. 2001). Indus and Ganges dolphins are obligate freshwa-
ter cetaceans that occur in several thousand kilometers of river
from the foothills of the Himalayas to the river deltas, and
occasionally in the ocean within a river’sfreshwaterplume
Electronic supplementary material The online version of this article
(doi:10.1007/s10914-014-9265-6) contains supplementary material,
which is available to authorized users.
G. T. Braulik :V. O d on :V. Islas-Villanueva :J. A. Graves
University of St. Andrews, St. Andrews, Fife, KY16 8LB, UK
G. T. Braulik (*)
Pakistan Wetlands Programme/WWF-Pakistan, House 62, Street 5,
Sector F10/1, Islamabad, Pakistan
e-mail: gillbraulik@downstream.vg
R. Barnett :A. R. Hoelzel
School of Biological and Biomedical Sciences, Durham University,
Durham DH1 3LE, UK
V. Islas-Villanueva
Instituto de Ciencias del Mar y Limnologia, Universidad Nacional
Autonoma de Mexico (UNAM), Circuito exterior s/n, Ciudad
Universitaria, 04510 México, D.F., México
J Mammal Evol
DOI 10.1007/s10914-014-9265-6
(Smith et al. 2006). Both subspecies are listed as endangered
on the IUCN Red List and they face numerous direct and
indirect threats. Despite this threatened status, the taxonomic
relationship between the two dolphin populations has never
been comprehensively evaluated, although it has been the
subject of conjecture (Pilleri et al. 1982; Reeves and
Brownell 1989;Rice1998).
The Ganges River dolphin was described in two separate
accounts both written in 1801 (Lebeck 1801; Roxburgh 1801).
The Indus River dolphin was described later, as Platanista indi
by Blyth (1859), and in 1976 was renamed P. minor due to an
overlooked earlier description (Owen 1853).
Anderson (1879) compared the external body morphology,
skull morphology, and skeletons of Indus and Ganges River
dolphins and concluded that evidence did not support the
recognition of different species, only that males were smaller
than females and that individuals exhibited geographic size
variation. Based on these conclusions, from the 1880s until
the 1970s, Indus and Ganges River dolphins were considered
to be subspecies. In the 1970s a number of comparative
studies were conducted concluding that Indus and Ganges
River dolphins possessed significant differences in the nasal
crests on the skull (Pilleri and Gihr 1971), between the sixth
and seventh cervical vertebrae (Pilleri and Gihr 1976), the
composition of blood proteins (De Monte and Pilleri 1979),
lipids in the blubber (Pilleri 1971), and in the length of the tail
(Kasuya 1972). Based on this evidence, they were referred to
as separate species until the late 1990s (Pilleri et al. 1982).
Limited sampling and lack of statistical analyses in the
previous studies led Rice (1998) to conclude that Platanista
comprised a single species with two subspecies, which is the
classification currently recognized by the Society for Marine
Mammalogy (Committee on Taxonomy 2013). It is likely that
the recent change to sub-specific status for the Indus and
Ganges dolphins has reduced their conservation priority
(Reeves et al. 2004). The numerous changes in taxonomy
imply that new information was available on which to base
these decisions; however, the comparative studies conducted
have all been very limited and do not provide sufficient weight
of evidence to defend any taxonomic classification.
Especially for endangered species like the South Asian
river dolphins, alpha taxonomy and recognition of evolution-
arily significant units (ESUs) within species are important
components of conservation strategies towards the preserva-
tion of genetic diversity and evolutionary potential. As incor-
rect taxonomic classifications often have serious conse-
quences for wildlife conservation, clarification of the system-
atics of Platanista has been listed as amongst the highest
priority of all cetaceans (Reeves et al. 2004). Here we present
the first quantitative evaluation of the phylogenetic relation-
ship between the Indus and Ganges River dolphins and the
estimated time of their divergence.
Methods
Very few Platanista tissue samples are available worldwide
and it is not possible to biopsy live South Asian river dolphins
Fig. 1 The Indus, Ganges, Brahmaputra, and Karnaphuli River systems in South Asia and the geographic origin of samples used in this study. Thick
black lines depict the approximate current range of Platanista based on descriptions in Sinha et al. (2010)
J Mammal Evol
to obtain additional samples because they surface unpredict-
ably and rapidly, and do not approach boats. Therefore, the
majority of samples used in this study were DNA extracted
from Platanista skeletal specimens stored in museums. Two
fresh tissue samples were also available.
A total of 65 Platanista samples were collected: 39 Indus
River dolphins, 18 Ganges River dolphins, and eight where
the river of origin was unknown (Table A, Supplementary
Information for details of year, origin, tissue, and museum).
The museum specimens were dated from two time periods,
either the 1800s or the 1970s; therefore, DNA was either
approximately 30 years old (43 samples) or was between
110 and 160 years old (15 samples). Five samples had no
information on collection date but are likely to be pre-1900
because newer specimens generally had more associated in-
formation. Prior to 1947, Indus and Ganges River dolphins
both occurred in British India. Museum specimens collected
prior to 1947, with an origin described only as India could
therefore have originated from any of the subcontinent’srivers
so these samples were classified as of unknown provenance.
Prior to sampling museum specimens, the work-area and
tools were cleaned with bleach to remove contaminating
particles. An electric drill with 3 mm bit was operated at low
speed (6,000 rpm) to extract bone powder from the densest
part of the bone. A new drill bit was used on each specimen to
prevent cross-contamination. In addition to bone powder,
dried soft tissue was collected from 14 specimens using a
sterile scalpel. Samples were stored in labelled, sealed, plastic
bags and frozen.
DNA Extraction and Amplification
DNA extractions from the samples were performed in the
ancient DNA laboratory at Durham University. All materials
and work surfaces were bleached before use with a 10 %
dilution of sodium hypochlorite and the workspace was UV
irradiated overnight. Samples of Platanista bone fragments
and preserved tissues were manually reduced to bone powder
or macerated tissue, and incubated overnight on a rotator at
55 °C in 500 μl of extraction buffer (1 M EDTA, 15 mM Tris,
pH8.0, 1 %w/v SDS) with 8 μl of Proteinase K (0.3 mg.ml–1).
Digested samples were then extracted using the QIAquick
PCR purification method of Yang et al. (1998). Final eluates
of DNA were collected in 50 μlofTEbuffer(1mMEDTA,
10 mM Tris, pH8.0) and stored at −20 °C. Negative extraction
controls were performed in parallel at a ratio of approximately
1:7. Partial (458 bp) control region sequences were amplified
in two overlapping fragments of 178 and 302 bp. PCR reac-
tions were 2μL of DNA extract in a 25 μl volume with Hi-
Fidelity Platinum Taq (Invitrogen, UK). Final PCR conditions
were: 2.5 mM dNTPs, 1 mM MgCl2, 1U Taq, and 200 μMof
each primer (Table B), plus buffer. The PCR program was an
initial cycle of 95 °C for 5 min, 45 cycles of 95 °C for 45 s,
58 °C for Frag 1 or 52 °C for Frag 2 for 45 s, 68 °C for 45 s and
68 °C for 5 min. PCR products were purified using the
QIAquick purification kit.
Following the analysis of sequences generated by Durham
University, it was clear that variability was extremely low, and
that longer control region fragments were better able to illus-
trate differences between dolphin populations. We therefore
conducted supplementary laboratory work at the University of
St. Andrews to attempt to boost the sample size of Indus River
dolphin sequences to include the complete control region
(856 bp) and the cytochrome b gene (1,111 bp) to add to
several Ganges dolphin sequences of these genes already
available on GenBank. DNA extraction at the University of
St. Andrews was conducted using the standard salt-saturated
extraction technique (Sunnucks and Hales 1996)onlyontwo
fresh tissue samples newly obtained (see Table A). The com-
plete control region and cytochrome b genes were each am-
plified and sequenced in two overlapping fragments (Fig. A)
as follows. PCRs used the MyTaq Red mix inclusive of dye
from Bioline. 12.5 μL of the enzyme mix was supplemented
with primers (working concentration of 0.25 μM) and ddH
2
O
to a final volume of 25 μL. The general PCR program was:
initial cycle of 95 °C for 5 min, 35 cycles of 95 °C for 45 s,
58 °C for 1 min, 72 °C for 1 min, and 10 min at 72 °C.
DNA was quantified and checked for quality with a
Nanodrop 1,000 spectrophotometer from Thermo Scientific,
working aliquotes of 50 ng/μL were created for PCR ampli-
fication. The control region and cytochrome b fragments were
amplified from 2 μL of DNA extract for 35 cycles. Fragments
were gel purified using the WIZARD SV gel and PCR clean-
up system from Promega, and eluted in 50 μl of ddH
2
O,
concentrated under vacuum at 45 °C for 40 min and re-
suspended in 10 μL of ultra-pure water. For ten samples, the
amount of recovered purified DNA was too low to produce a
readable chromatogram. The PCR was therefore repeated
using 5 μL of DNA extract, amplified for an initial 25 cycles.
The PCR fragments were gel excised with disposable scalpels,
purified and concentrated as above, and subjected to a second
amplification for 15 cycles. Negative controls were included
with all amplifications and a positive control from a soft tissue
extraction. The purified PCR products were sequenced in both
directions using ABI BigDye Terminator. The methods used
are summarized in Fig. A, and primers are detailed in Table B
(Supplementary Information).
Genetic Analyses
Sequences were aligned using the software ClustalW 2.1
(Larkin et al. 2007), and were checked by eye to ensure
optimal alignment. Control region analysis was conducted
on partial sections 458 bp in length, and additional analyses
conducted on a subset of longer control region sequences of
856 bp; all sequences were trimmed using the software
J Mammal Evol
MEGA 5.05 (Tamura et al. 2011). The number of haplotypes,
haplotype (h) and nucleotide diversity (π), and the number
and type of single nucleotide polymorphisms were assessed
using ARLEQUIN 3.5 (Excoffier and Lischer 2010). Haplo-
type networks were constructed using a median-joining algo-
rithm in the software NETWORK 4.6 (Bandelt et al. 1999).
Neutrality tests were applied using Fu’s F and Tajima’s D tests
in ARLEQUIN. The degree of differentiation between Indus
and Ganges River dolphins was evaluated using F
ST
and Φ
ST
scores, with the Tamura-Nei genetic distance model (as the
HKY model is not available in Arlequin and Tamura-Nei was
also highly ranked in jModelTest, see below) and 1,000 per-
mutations to create p-values. Nei’s pairwise distances were
compared between and within populations (Nei and Li 1979),
and an exact test of population differentiation was performed
with 10,000 Markov chain steps.
Phylogenetic trees were constructed using Maximum Par-
simony (MP) and Maximum Likelihood (ML) methods in
MEGA 5.05 and by Bayesian Markov Chain Monte Carlo
(MCMC) simulations using MrBayes 3.2 (Ronquist et al.
2012). Only six samples, all from the Indus, yielded both the
control region and cytochrome b, so we were unable to test the
concatenated DNA. The six haplotypes identified from the
complete Platanista mtDNA control region were used to
create the trees, and four allied species used as outgroups:
Baird’s beaked whale (Berardius bairdii) NC_005274, north-
ern bottlenose whale (Hyperoodon ampullatus) NC_005273,
pygmy sperm whale (Kogia breviceps) NC_005272 (Arnason
et al. 2004), and sperm whale (Physeter macrocephalus)
NC_002503 (Arnason et al. 2000). ML and MrBayes used
the Hasegawa-Kishino-Yano (HKY) model of nucleotide sub-
stitution identified by jModelTest 2 (Darriba et al. 2012)asthe
most appropriate model based on Akaike’s Information Crite-
rion (AIC) values. The Close-Neighbor-Interchange algorithm
was used to construct the MP tree. The Bayesian analysis used
the default priors on branch lengths, rate parameters, and tree
topology with four chains of 3,000,000 generations sampled
every 5,000 generations and no hot chains. The first 90,000
generations were discarded as burn-in. The 60 % bootstrap
consensus trees inferred from 5,000 replicates were generated
for each of the methods and the resulting topology, bootstrap
values, and posterior probabilities compared.
The time of divergence between lineages was estimated
using a strict molecular clock and an HKY model of substitu-
tion in the Bayesian phylogenetic software BEAST 1.6.1
(Drummond and Rambaut 2007). One complete mtDNA con-
trol region sequence from the Indus and one from the Ganges
representing the two most closely allied haplotypes (Indus
HAP1 and Ganges HAP4) were included in the model. Coa-
lescent tree priors are most suitable for describing the rela-
tionships between individuals in the same population/species,
whereas a Yule tree prior is most appropriate for between-
species comparisons. Since we were exploring whether there
are potentially species-level differences between two popula-
tions, we compared the results generated when using each
prior. Divergence times can be calibrated either by specifying
a substitution rate or by calibrating one of the nodes of the tree
based on fossil evidence or other information. Cetaceans have
been shown to have considerable substitution rate heteroge-
neity among lineages and over time; for example, the
mysticetes have one of the slowest mtDNA substitution rates
of all mammals, while odontocetes were estimated to have
rates three to four times as fast, and a clade comprising Iniidae
and Pontoporiidae almost twice as fast again (Dornburg et al.
2012). Given this, and that Platanistidae is not closely related
to any other cetacean family, we did not find it possible to
reliably or defensibly select a clock rate from another cetacean
to calibrate our phylogenetic tree. Platanistids split from other
cetaceans very early, the affinity to, or with, earlier fossils is
not clear, and the fossil record is patchy (Barnes et al. 1985).
For this reason, fossil dates were not used for calibration, and
we instead used the divergence time between the Platanistidae
and a clade including Ziphiidae (Clade G: 28.77 MYago, log-
normal SD=0.08), which was estimated with high confidence
(1.0 95 % highest posterior density (HPD)) in a time-
calibrated phylogeny of whales (Xiong et al. 2009). The
Ziphiidae (AJ554056) and Platanista (AJ55408) sequences
used by Xiong et al. (2009) in their phylogeny were also used
to calibrate the current model. The models were run for
50,000,000 MCMC steps, and were sampled every 1,000
steps, which was sufficient to ensure convergence on a sta-
tionary distribution for each parameter. Visual inspection of
traces supported the removal of the first 10 % of the MCMC
chains as burn-in, and convergence statistics were monitored
by effective samples sizes. A sperm whale sequence was used
as the outgroup (GenBank: NC_002503).
Results
mtDNA Control Region Sequences
A 458 bp section of the mtDNA control region was success-
fully amplified and sequenced from 28 samples, comprising
25 from the Indus and three from the Ganges. These were
combined with 14 sequences on GenBank (Cassens et al.
2000; Hamilton et al. 2001;Yangetal.2002; Arnason et al.
2004; Verma et al. 2004) (details in Supplementary Information,
Tab le A) to give a total sample of 42 (26 Indus and 16 Ganges)
for the analysis. The sample source, species, and location of
origin of GenBank samples were verified by reference back to
the published paper or by direct correspondence with the author,
and none were duplicates of the museum specimens sampled.
Only the origin of the GenBank P. minor sample AJ554058
(Arnason et al. 2004) could not be traced. Despite numerous
attempts, no usable DNA was extracted from any specimen
J Mammal Evol
collected in the 19th century (most of which were Ganges River
dolphins). Eleven Indus dolphin samples were successfully se-
quenced by both laboratories; in all cases the sequences obtained
were identical, which argues strongly for their authenticity. Fig. 1
illustrates the geographic location of origin of the sequences
obtained; all Indus sequences originated in Sindh, which is the
remaining high density area for that subspecies, while Ganges
dolphin sequences were obtained from four locations across the
majority of their current area of occurrence.
All partial control region sequences from the Indus River
dolphin were the same haplotype. Within the Ganges River
dolphin sample, there were two haplotypes, separated by a
single transition (Table 2). Although higher than the Indus
dolphin, Ganges River dolphin nucleotide and gene diversities
were both very low (Table 1). None of the haplotypes were
shared between Indus and Ganges River dolphins, and there
were two fixed transitional differences separating them.
As the analysis showed such low genetic variability, to
investigate whether the analyzed 458 bp portion was more
highly conserved than other parts of the control region, the
above analyses were repeated on the 31 samples for which the
entire 856 bp control region had been sequenced (17 from this
study and 14 from GenBank; 18 Indus and 13 Ganges) and the
results compared. These longer sequences had similar diversity
indices to the partial sequences, but contributed three additional
haplotypes, two from the Indus and one from the Ganges River
dolphin giving a total of six, three in each population (Tables 1
and 2). None of the six haplotypes were shared between Indus
and Ganges dolphins, and there were five fixed differences
between them, comprising three transitions, one transversion,
and one insertion-deletion. The median-joining haplotype net-
work from the samples was very simple, and reflected the low
variability and few haplotypes (Gachal 2001) found within each
population (Fig. 2). Haplotype (HAP-4) was shared by dolphins
that originated from the Ganges River at Patna, and also by two
specimens collected from the separate Karnaphuli River system
approximately 1,000 km away in eastern Bangladesh. New
haplotypes from this study were submitted to Genbank as
accession numbers KJ629309-313 for the control region
(HAP-3 was derived from AJ554058 and does not have a
new accession number), and KJ620052-55 for cytochrome b.
Taj ima ’s D test was not significant for either the Indus (D=
−1.165, p=0.135) or the Ganges (D=1.550, p=0.959) sam-
ples; however, a significant result was obtained for Fu’sF
s
test
in the Indus dolphin population (Indus: F
s
=-1.744, p<0.02;
Ganges: F
s
=0.774, p=0.631), which suggests recent popula-
tion demographic expansion.
Cytochrome b Sequences
We sequenced 1,111 bp of the cytochrome b gene for 13 Indus
dolphins. Four shorter sequences were obtained as unpub-
lished data (Gachal 2001) and nineteen Platanista cytochrome
b sequences were available on GenBank (one Indus; 18 Gan-
ges), to give a total of 36, comprising 18 Indus dolphins and
18 Ganges dolphins. The GenBank sequences were of vari-
able length requiring the entire dataset to be trimmed to a
541 bp portion that was present in 29 of the 36 samples (12
Indus; 17 Ganges), which were then used for the analysis. As
for the control region, gene and nucleotide diversities were
very low, and higher in the Ganges River dolphin than the
Indus; gene diversity: 0.559±0.083 versus 0.485 ± 0.106; and
nucleotide diversity: 0.0014 ± 0.0012 versus 0.0009±0.0009,
for the Ganges and Indus, respectively. There were three
polymorphic sites, all transitions, and four haplotypes, one
unique to the Indus, two unique to the Ganges, and one shared
by both (Table 3). All haplotypes were separated by a single
transition.
Population Differentiation, Phylogeny, and Divergence
The fixation indices between Indus and Ganges dolphins were
all very high, and clearly significant; Φ
ST
for the partial control
region sequences was 0.932 (p<0.001) and F
ST
was 0.843
(p<0.0001). Furthermore, for the control region, the exact test
of differentiation between the two rivers was significant
(p<0.0001), and the average number of pairwise differences
were significantly greater when comparing between popula-
tions than within populations (D=5.269, p<0.0001). Based on
the complete control region sequences, net divergence between
Indus and Ganges dolphins was 0.58 %, compared to 0.23 %
within each population. The phylogenetic reconstructions
Tabl e 1 Nucleotide and haplotype diversity in mtDNA control region samples from the Indus and Ganges-Brahmaputra Rivers
Population Nucleotide diversity Haplotype diversity
Partial 458 bp control region
portion n=42
Complete 858 bp control
region n=31
Partial 458 bp control region
portion n=42
Complete 858 bp control
region n=31
Indus dolphin Partial, n=26
Complete, n=18
0± 0 0.0003±0.0004 0±0 0.216±0.124
Ganges dolphin Partial, n=16
Complete, n=13
0.0009±0.0010 0.0012±0.0009 0.400± 0.114 0.641±0.097
Both combined 0.0025±0.0018 0.0019±0.0014 0.539±0.062 0.682±0.070
J Mammal Evol
showed separation between Indus and Ganges dolphins
that was supported with 90 % posterior probability in the
Bayesian tree, and in 95 and 84 % of the bootstrap
replicates in the ML and MP trees, respectively (Fig. 3).
The modelling of divergence time estimated that the two
populations have been separated for 0.549 MY (0.125–
1.049 MY 95%HPD) with a clock rate of 5.4×10
−3
substitutions/site/MY when using the Yule speciation pri-
or, and for 0.555 MY (0.133–1.059 MY 95%HPD) with a
clock rate of 5.3×10
−3
substitutions/site/MY when the
coalescent constant population size prior was used.
Discussion
Lack of Variability
The control region is a highly variable part of the mammalian
mitochondrial genome, and the total absence of variability in
458 bp of the partial Indus dolphin control region is unusual.
Amongst cetaceans, control region homogeneity has only
previously been recorded in 322 bp of the vaquita (Phocoena
sinus) (Rosel and Rojas-Bracho 1999) and 206 bp of the
Maui’sdolphin(Cephalorhychus hectori maui)(Pichlerand
Baker 2000). These are both critically endangered species/
subspecies with restricted ranges and small populations that
are declining due to human activities. The low diversity in
Maui’sdolphinwasattributedtorecentpopulationdepletion
caused by fisheries bycatch (Pichler and Baker 2000), and
genetic homogeneity in the vaquita was proposed to be due to
historically low population size and a founder event in its
origin (Taylor and Rojas-Bracho 1999). Homogeneity in the
Indus River dolphin sequences and the significant expansion
signal from Fu’s Fs test both suggest that the population has
expanded after passing through a bottleneck in the past. In the
1970s, the Indus dolphin declined to perhaps just a few
hundred individuals due to a century of intense hunting; since
Table 2 Haplotypes identified in the entire (858 bp) mitochondrial
control region, and the less resolved, partial (458 bp) control region
sequences (shaded grey), along with sample size, and river of origin,
for two geographically isolated populations of South Asian river
dolphins. The position in the sequence where the substitution occurred
is numbered in the header, nucleotides underlined and in bold represent
fixed differences between the Indus and Ganges River dolphins
Population
No. of
sequences* 41 71 123 140 147 297 418 633 704
Indus 26 CTT
Hap-1 Indus 16 ACIndel TTCTT Indel
Hap-2 Indus 1 ACIndel TACTT Indel
Hap-3 Indus 1 ACIndel TTCTTC
Ganges 12 TCT
Hap-4 Ganges 7 GCCATTCT Indel
Hap-5 Ganges 4 GTCATTCT Indel
Ganges 4 TCC
Hap-6 Ganges 2 GTCATTCC Indel
*Number of partial and complete sequences are not additive
Tab l e 3 Haplotypes and variable sites in a 541 bp portion of the
Platanista gangetica cytochrome b gene
Position 307 367 426 Number
Indus Ganges
Hap-1 G A T –1
Hap-2 A G C –10
Hap-3 G A C 4 –
Hap-4 G G C 8 6
J Mammal Evol
our samples are from the 1970s a post-bottleneck expansion
likely predated that time.
In addition to the partial control region uniformity in the
Indus dolphin, the entire Platanista genus has extremely low
mtDNA genetic variation. This may be because the species
occurs in restricted habitat and is naturally not abundant.
Although there is a perceived link between a lack of genetic
variation and an increased risk of extinction for small popu-
lations (Gilpin and Soule 1986), low mitochondrial variability
in this species is not the biggest issue it faces, since there is
now immense human pressure on its habitat.
Divergence
The Indus and Ganges River dolphin populations appear to
have been reproductively isolated since sharing a common
ancestor approximately 0.55 MY ago. This calculation of
divergence is very similar to the 0.51 MY (95 %pp: 0.14–
1.02) estimated by McGowen et al. (2009) using Indus and
Ganges River dolphin cytochrome b sequences. Ho et al.
(2008) demonstrated that divergence estimates from molecu-
lar ecological studies can be altered by the choice of calibra-
tion points, with deeper, external calibration points, such as
that used in our study, sometimes leading to overestimates of
times to divergence. For example, for bowhead whales, clock
rates increased by almost an order of magnitude when internal
calibration was used (Ho et al. 2008). The clock rate estimated
by our Bayesian analysis was quite slow, and similar to that
estimated for mysticetes by Dornburg et al. (2012). This may
be accurate, or it may be underestimated because of our
calibration point. Although internal calibration points may
be more accurate, they also have a much wider degree of
uncertainty, and in this case no suitable internal calibration
points were available. The node we used to calibrate our
divergence time was 28.77 MY (Xiong et al. 2009); this node
has also been estimated in other phylogenetic studies as 32.43
Fig. 3 Maximum parsimony (MP) bootstrap consensus tree with
branches corresponding to partitions reproduced in less than 60 % of
the bootstrap replicates collapsed. Numbers next to the branches refer to
the percentage of replicate trees in which the taxa clustered together in the
bootstrap test. The node between the Indus and Ganges sequences was
supported in 84 % of the MP and 95 % of the Maximum likelihood (ML)
trees and with 90 % posterior probability in the Bayesian tree
Fig. 2 Median-joining network based on complete mtDNA control
region haplotypes for the cetacean family Platanistidae. Circle size is
proportional to the number of individuals representing that haplotype
(HAP-2 and 3 represent single animals) and branch lengths are
approximately proportional to the number of mutations. Transitions are
represented by thin bars perpendicular to each branch, transversions by
thick grey bars, and insertion-deletions by thick black bars. Light grey =
Ganges River dolphin, dark grey = Indus River dolphin
J Mammal Evol
MY (27.92–37.07 MY) (McGowen et al. 2009), 30.50 MY±
1.3 (Arnason et al. 2004), 28.9 MY± 4.9 (Nikaido et al. 2001),
and based on the fossil record as 23 MY (Hamilton et al.
2001). To test the sensitivity of our estimate of divergence to
changes in the calibration point, the Bayesian phylogenetic
analysis was duplicated using the longest (32.43 MY) and
shortest (23 MY) node divergence estimates. When the lon-
gest node estimate was used, Platanista divergence was esti-
mated at 0.62 MY (0.14–1.17 MY), and the shortest calibra-
tion node estimate generated a divergence date of 0.44 MY
(95%pp: 0.10–0.83 MY) demonstrating a consistency in our
estimates. It is important to note, however that the 95 %
posterior probability on all of these estimates is quite wide.
Paleo-Drainage Patterns and River Capture
There is overwhelming evidence that the Indus, Brahmaputra,
and Ganges Rivers, existed as separate river systems soon
after the original India-Asia collision (~55–45 Million Years
(MY) ago) and even before large-scale elevation of the Hima-
layan mountains (Clift and Blusztajn 2005). Analysis of the
Bengal fan shows that sedimentation accelerated in the Mio-
cene, and that there has been no major shift in either the area
being drained or the location of the river mouth since that time
(Burbank et al. 1993). Although the Indus and Ganges-
Brahmaputra have been separate rivers for many millions of
years, seismic reflection data and neodymium isotopes in the
Indus fan clearly demonstrated that the five major tributaries
of the Indus, which together contribute flow approximately
equal to the Indus itself (Fig. 1), were formerly connected to
the Ganges River and were progressively captured by the
Indus sometime after 5 MY ago (Clift and Blusztajn 2005).
Other than the transfer sometime after 50,000 years ago of the
Yamuna River from the Indus to the Ganges (Clift et al. 2012),
the timing of specific capture events cannot be reconstructed
from the terrestrial evidence alone and in many cases remains
unclear or imprecise (Clift and Blusztajn 2005).
Within river systems there are often strong relationships
between the present distribution of aquatic species and histor-
ical drainage patterns (Waters et al. 2001). This is because for
exclusively aquatic animals, such as freshwater fish or obli-
gate river dolphins, both the terrestrial environment and the
ocean represent formidable barriers to dispersal between
drainages. Inter-drainage transfer and cladogenesis for
freshwater-limited taxa often occurs through river capture,
the process through which stream sections are displaced from
one catchment to another (Burridge et al. 2006). Our molec-
ular data indicate that there was a significant drainage change
around 0.55 MY ago that allowed for freshwater dolphins to
transfer from one river system to the other. Because of their
concentration in the main stem of large rivers, movement of
dolphins between drainages would have required the capture
of a major river such as one of the largest Indus or Ganges
tributaries (Fig. 1). Our results corroborate the documented
drainage re-organization that occurred on the Indo-Gangetic
plains, and the theory that beginning around 5 MY ago and
completed by 0.3 MY ago the major tributaries of the Ganges
in Punjab were gradually re-routed into the Indus (Clift and
Blusztajn 2005). We hypothesize that the South Asian river
dolphins arose in the larger Ganges-Brahmaputra River sys-
tem and later colonized the adjacent Indus River system,
around 0.5–0.6 MY ago when a large lowland tributary
inhabited by Platanista was captured.
River Dolphin Speciation and Divergence
Both the Biological Species Concept (BSC) and the Phyloge-
netic Species Concept (PSC) are relevant, and have been used
effectively, for cetacean taxonomy. Reeves et al. (2004)con-
cluded that that these different approaches to species delimi-
tation should be employed in a flexible and pragmatic way,
with the basic aim of using multiple lines of evidence as
proxies to demonstrate irreversible divergence. The most
commonly applied criteria used to define phylogenetic species
is the necessity of reciprocal monophyly for mtDNA alleles,
which means that all DNA lineages must share a more recent
common ancestor with each other than with lineages from
other ESUs (Moritz 1994). The strength of this approach is
that it avoids the issue of ‘how much divergence is enough?’
that plagues quantitative criteria such as allele frequency
divergence and genetic distance, and it considers the pattern
rather than the extent of sequence divergence (Moritz 1994). It
has also been shown that the requirement of monophyly at
neutral loci can be too conservative for the purposes of tax-
onomy or conservation, because recent speciation events such
as the divergence of the finless porpoises (Neophocaena
phocaenoides and Neophocaena asiaeorientalis) may not be
detected due to the lack of fixed molecular differences be-
tween recently derived species (Wang et al. 2008). This is not
the case for the Indus and Ganges dolphins where reciprocal
monophyly in the mtDNA control region was clearly present,
there were no shared haplotypes, and the requirements of the
PSC were satisfied. The very high F
ST
statistics (Φ
ST
=0.932;
F
ST
=0.843) reflect the low diversity and fixed differences.
Dispersal between river systems and contemporary ge-
netic exchange is deemed very unlikely since the popula-
tions occur in river systems separated by several hundred
kilometers of land at their closest point, and although these
dolphins are occasionally sighted in brackish water, their
dispersal between river systems through the ocean would
involve a highly improbable journey through exposed
saline waters, of at least 4,600 km around the Indian
peninsula.
One limitation of this study is that sampling locations were
separated by considerable distances and there were no samples
from close to the Indus-Ganges drainage divide (Fig. 1).
J Mammal Evol
Although there was no evidence of genetic structure within the
Ganges dolphin samples, more comprehensive sampling
would help to exclude the possibility that some of the differ-
ences observed were due to isolation by distance. The Indus
and Ganges dolphins are geographically isolated, and obvi-
ously on an independent evolutionary trajectory; however, to
adequatelydemonstrate the irreversible divergence required to
be able to support their consideration as separate species, there
is a need to build on this study with comparative investiga-
tions of skeletal or external morphology or with additional
genetic data. The material we could access was extremely
difficult to obtain, and we believe it would have been too
degraded for the successful amplification of nuclear loci by
conventional means. Future work may require DNA capture
from museum specimens and next generation sequencing
technologies if sufficient amounts of better quality material
cannot be compiled. The results from this study raise the
possibility that additional comparative investigations may
allow recognition of these isolated populations as distinct taxa.
Given the rapid decline in range of both the Indus and Ganges
dolphins over the last century and the ongoing degradation of
their habitats, resolving the taxonomic relationship between
these animals is of high importance for their conservation.
Acknowledgments For facilitating access to museum specimens we
would like to thank Doris Moerike: Staatliches Museum für Naturkunde
Stuttgart, Richard Sabin: British Museum of Natural History, Martin
Milner: Bell-Pettigrew Museum, University of St. Andrews, Mathew
Lowe and Robert Asher: Cambridge University Museum of Zoology,
Andrew Kitchener and Jerry Herman: National Museums of Scotland,
Edinburgh, Uzma Khan: WWF-Pakistan, Kelly Robertson: National
Marine Fisheries Service USA, Tom Jefferson and Bill Perrin. We thank
Tanya Sneddon for lab assistance at St Andrews, and Darren Parker for
guidance on MrBayes. The work was funded by the Whale and Dolphin
Conservation Society and the US Marine Mammal Commission. We
thank Simon Northridge, Phil Hammond, and Peter Clift for reviewing
earlier versions of the manuscript.
References
Anderson J (1879) Anatomical and Zoological Researches: Comprising
an Account of the Zoological Results of the Two Expeditions to
Western Yunnan in 1868 and 1875 and a Monograph of the Two
Cetacean genera Platanista and Orcella. Bernard Quaritch,
Piccadilly, London
Arnason U, Gullberg A, Gretarsdottir S, Ursing B, Janke A (2000) The
mitochondrial genome of the sperm whale and a new molecular
reference for estimating eutherian divergence dates. J Mol Evol
50(6): 569–578
Arnason U, Gullberg A, Janke A (2004) Mitogenomic analyses provide
new insights into cetacean origin and evolution. Gene 333: 27–34
Bandelt H-J, Forster P, Röhl A (1999) Median-joining networks for
inferring intraspecific phylogenies. Mol Biol Evol 16: 37–48
Barnes LG, Doming DP, Ray CE (1985) The status of studies on fossil
marine mammals. Mar Mammal Sci 1: 15–53
Blyth E (1859) On the rorqual of the Indian Ocean, with notices of other
cetals, and of the sirenia or marine pachyderms. J Asiatic Soc Bengal
28: 481–498
Burbank DW, Derry LA, France-Lanord C (1993) Reduced Himalayan
sediment production 8 Myr ago despite an intensified monsoon.
Nature 364: 48–50
Burridge CP, Craw D, Waters JM (2006) An empirical test of freshwater
vicariance via river capture. Mol Ecol 16: 1883–1895
Cassens I, Vicario S, Waddell GV, Balchowsky H, Van Belle D, Ding W,
Fan C, Mohan LRS, Simoes-Lopes PC, Bastida R, Meyer A,
Stanhope MJ, Milinkovitch MC (2000) Independent adaptation to
riverine habitats allowed survival of ancient cetacean lineages. Proc
Natl Acad Sci USA 97: (21) 11343–1134 7
Clift PD, Blusztajn J (2005) Reorganisation of the western Himalayan
river system after five million years ago. Nature 438: 1001–1003
Clift PD, Carter A, Giosan L, Durcan J, Duller GaT, Macklin MG,
Alizai A, Tabrez AR, Danish M, Vanlaningham S and Fuller
DQ (2012) U-Pb zircon dating evidence for a Pleistocene
Sarasvati River and capture of the Yamuna River. Geology
40(3): 211–214
Committee on Taxonomy (2013) List of marine mammal species and
subspecies. Society for Marine Mammalogy, www.
marinemammalscience.org, consulted on 26 Oct 2013
Darriba D, Taboada GL, Doallo R, Posada D ( 2012) jModelTest 2: more
models, new heuristics and parallel computing. Nature Methods 9:
772
De Monte T, Pilleri G (1979) Cetacean haematology 1: haemoglobin.
Investigations on Cetacea 10: 277–287
Dornburg A, Brandley MC, Mcgowen MR, Near TJ (2012)
Relaxed clocks and inferences of heterogeneous patterns of
nucleotide substitution and divergence time estimates across
whales and wolphins (Mammalia: Cetacea). Mol Biol Evol
29(2): 721–736
Drummond AJ, Rambaut A (2007) BEAST: bayesian evolutionary anal-
ysis by sampling trees. BMC Evolutionary Biology 7: 214
Excoffier L, Lischer HEL (2010) Arlequin suite ver 3.5: a new series of
programs to perform population genetics analyses under Linux and
Windows. Mol Ecol Resources 10: 564–567
Gachal GS (2001) Aspects of the environmental ecology and an intro-
duction to the molecular genetics of the Indus River dolphin
(Platanista minor). PhD Thesis, School of Bioscience, Cardiff
University, Cardiff, UK
Gilpin ME, Soule ME (1986) Minimum viable populations: processes of
species extinction. In: Soule ME (ed) Conservation Biology: The
Science of Scarcity and Diversity. Sinauer Associates Inc.,
Sunderland, Massachusetts
Hamilton H, Caballero S, Collins AG, Brownell JRL (2001) Evolution of
river dolphins. Proc R Soc London 268: 549–556
Ho SYW, Saarma U, Barnett R, Haile J, Shapiro B (2008) The effect of
inappropriate calibration: three case studies in molecular ecology.
PLoS ONE 3(2): e1615
Kasuya T (1972) Some informations on the growth of the Ganges dolphin
with a comment on the Indus dolphin. Sci Rep Whales Res Inst 24:
87–108
Larkin MA, Blackshields G, Brown NP, Chenna R, Mcgettigan PA,
Mcwilliam H, Valentin F, Wallace IM, Wilm A, Lopez R,
Thompson JD, Gibson TJ, Higgins DG (2007) Clustal W and
Clustal X version 2.0. Bioinformatics 23: 2947–2948
Lebeck HHJ (1801) Delphinus gangeticus beschrieben von Heinrich
Julius Lebeck zu Trankenbar. Der Gesellschaft Naturforschender
Freunde zu Berlin Neue Schriften 3: 280–282
McGowen MR, Spaulding M, Gatesy J (2009) Divergence date estima-
tion and a comprehensive molecular tree of extant cetaceans. Mol
Phylogen Evol 53: 891–906
Moritz C (1994) Defining ‘evolutionary significant units’for conserva-
tion.TrendsEcolEvol9:373–375
J Mammal Evol
Nei M, Li WH (1979) Mathematical model for studying genetic variation
in terms of restriction endonucleases. Proc Natl Acad Sci USA 76:
5269–5273
Nikaido M, Matsuno F, Hamilton H, Brownell JRL, Cao Y, Ding W,
Zuoyan Z, Shedlock AM, Fordyce RE, Hasegawa M, Okado N
(2001) Retroposon analysis of major cetacean lineages: the mono-
phyly of toothed whales and the paraphyly of river dolphins. Proc
Natl Acad Sci USA 98(13): 7384–7389
Owen (1853) Descriptive Catalogue of the Osteological Series Contained
in the Museum of the Royal College of Surgeons. Volume II.
Mammalia Placentalia. Taylor and Francis, London
Pichler FB, Baker CS (2000) Loss of genetic diversity in the endemic
Hector’s dolphin due to fisheries-related mortality. Proc R Soc
London B 267: 97–102
Pilleri G (1971) Preliminary analysis of the lipids present in the blubber of
Platanista indi and gangetica. Investigations on Cetacea 3: 51–52
Pilleri G, Gihr M (1971) Differences observed in the skulls of Platanista
gangetica [Roxburgh, 1801] and indi [Blyth, 1859]. Investigations
on Cetacea 3: 13–21
Pilleri G, Gihr M (1976) Osteological differences in the cervical vertebrae of
Platanista indi and gangetica. Investigations on Cetacea 7: 105–108
Pilleri G, Marcuzzi G, Pilleri O (1982) Speciation in the Platanistoidea,
systematic, zoogeographical and ecological observations on recent
species. Investigations on Cetacea 14: 15–46
Reeves RR, Brownell JRL (1989) Susu Platanista gangetica (Roxburgh,
1801) and Platanista minor Owen, 1853. In: Ridgway SH, Harrison
RJ (eds) Handbook of Marine Mammals Vol 4: The First Book of
Dolphins. Academic Press Ltd., London, pp 66–99
Reeves RR, Perrin WF, Taylor BL, Baker CS, Mesnick SL (eds) (2004)
Report of the workshop on the shortcomings of cetacean taxonomy
in relation to the needs ofconservation and management. NOAA, La
Jolla, California, 97 p
Rice D (1998) Marine mammals of the world: systematics and distribu-
tion. Soc Mar Mammalogy Spec Pub 4: 1–231
Ronquist F, Teslenko M, Van Der Mark P, Ayres DL, Darling A, Höhn S,
Larget B, Liu L, Suchard MA, Huelsenbeck JP (2012) MrBayes 3.2:
efficient Bayesianphylogenetic inference and model choice across a
large model space. Syst Biol 61(3): 539–542
Rosel PE, Rojas-Bracho L (1999) Mitochondrial DNA variation in the
critically endangered vaquita Phocoena sinus Norris and
MacFarland, 1958. Mar Mammal Sci 15(4): 990–1003
Roxburgh W (1801) An account of a new species of Delphinus,an
inhabitant of the Ganges. Asiatick Researches 7: 170–174
Sinha RK, Behera S, Choudhary BC (2010) Conservation action
plan for the Gangetic river dolphin 2012–2020. National
Ganga River Basin Authority, Ministry of Environment and
Forests, India, 38 p
Smith BD, Braulik G, Strindberg S, Ahmed B, Mansur R (2006)
Abundance of Irrawaddy dolphins (Orcaella brevirostris)and
Ganges River dolphin (Platanista gangetica gangetica) estimated
using concurrent counts made by independent teams in waterways
of the Sundarbans mangrove forest in Bangladesh. Mar Mammal Sci
22(2): 527–547
Steeman ME, Bebsgaard MB, Fordyce RE, Ho SYW, Rabosky DL,
Nielsen R, Rahbek C, Glenner H, Sørensen MV, Willerslev E
(2009) Radiation of extant cetaceans driven by restructuring of the
oceans. Syst Biol 58(6): 573–585
Sunnucks P, Hales DF (1996) Numerous transposed sequences of mito-
chondrial cytochrome oxidase I–II in aphids of the genus Sitobion
(Hemiptera: Aphididae). Mol Biol Evol 13: 510–524
Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S (2011)
MEGA5: Molecular evolutionary genetics analysis using maximum
likelihood, evolutionary distance, and maximum parsimony
methods. Mol Biol Evol 28: 2731–2739
Taylor BL, Rojas-Bracho L (1999) Examining the risk of inbreeding
depression in a naturally rare cetacean, the vaquita (Phocoena
sinus). Mar Mammal Sci 15(4): 1004–1028
Verma S, Sinha R, Singh L (2004) Phylogetic position of Platanista
gangetica: insights from the mitochondrial cytochrome b and nu-
clear interphotoreceptor retinoid-binding protein gene sequences.
MolPhylogenEvol33:280–288
Wang JY, Frasier TR, Yang SC, White BN (2008) Detecting recent
speciation events: the case of the finless porpoise (genus
Neophocaena). Heredity 101: 145–155
Waters JM, Craw D, Youngson JH, Wallis GP (2001) Genes meet
geology: fish phylogeographic pattern reflects ancient, rather than
modern, drainage connections. Evolution 55(9): 1844–1851
Xiong Y, Brandley MC, Xu S, Zhou K, Yang G (2009) Seven new dolphin
mitochondrial genomes and a time-calibrated phylogeny of whales.
BMC Evol Biol 9: (20) doi:10.1186/1471-2148-1189-1120
Yang DY, Eng B, Waye JS, Dudar JC, Saunders SR (1998) Technical
note: improved DNA extraction from ancient bones using silica-
based spin columns. Am Phys Anthropol 105: 539–543
Yang G, Zhou K, Ren W, Ji G, Liu S, Bastida R, Rivero L (2002)
Molecular systematics of river dolphins inferred from complete
mitochondrial cytochrome-b gene sequences. Mar Mammal Sci
18(1): 20–29
Zhou X, Xu S, Yang Y, Zhou K, Yang G (2011) Phylogenomic analyses
and improved resolution of Cetartiodactyla. Mol Phylogen Evol 61:
255–264
J Mammal Evol
