Cytochrome c oxidase subunit 1 barcode data of fish of the Nayband National Park in the Persian Gulf...

DNA BARCODING

Cytochrome c oxidase subunit 1 barcode data of fish of theNayband National Park in the Persian Gulf and analysisusing meta-data flag several cryptic species

HOSSEINALI ASGHARIAN,* HOMAYOUN HOSSEINZADEH SAHAFI,† ARIA ASHJA ARDALAN,‡

SHAHROKH SHEKARRIZ§ and ELAHE ELAHI*¶**

*Department Biotechnology, College of Science, University of Tehran, No 13, Shafiie Alley, Qods St., Enghelab St., 14155-6455,

Tehran, Iran, †Iranian Fisheries Research Organization, 14155-6116, Tehran, Iran, ‡Faculty of Marine Science and Technology,

Islamic Azad University, North Branch of Tehran, 1987973133, Tehran, Iran, §Department of Biology, Islamic Azad University,

Qods Branch of Tehran, Qods Sq., Km 20 Tehran-Karaj Old Road, 374-37515, Tehran, Iran, ¶School of Biology, College of Science,

University of Tehran, Qods St., Enghelab St., 14155-6455, Tehran, Iran, **Center of Excellence in Biomathematics, School of Mathe-

matics, Statistics and Computer Science, College of Science, University of Tehran, Qods St., Enghelab St., 14155-6455, Tehran, Iran

Abstract

We provide cytochrome c oxidase subunit 1 (COI) barcode sequences of fishes of the Nayband National Park, Persian Gulf,

Iran. Industrial activities, ecological considerations and goals of The Fish Barcode of Life campaign make it crucial that fish

species residing in the park be identified. To the best of our knowledge, this is the first report of barcoding data on fishes

of the Persian Gulf. We examined 187 individuals representing 76 species, 56 genera and 32 families. The data flagged

potentially cryptic species of Gerres filamentosus and Plectorhinchus schotaf. 16S rDNA data on these species are provided.

Exclusion of these two potential cryptic species resulted in a mean COI intraspecific distance of 0.18%, and a mean inter- to

intraspecific divergence ratio of 66.7. There was no overlap between maximum Kimura 2-parameter distances among con-

specifics (1.66%) and minimum distance among congeneric species (6.19%). Barcodes shared among species were not

observed. Neighbour-joining analysis showed that most species formed cohesive sequence units with little variation.

Finally, the comparison of 16 selected species from this study with meta-data of conspecifics from Australia, India, China

and South Africa revealed high interregion divergences and potential existence of six cryptic species. Pairwise interregional

comparisons were more informative than global divergence assessments with regard to detection of cryptic variation. Our

analysis exemplifies optimal use of the expanding barcode data now becoming available.

Keywords: 16S rDNA, cytochrome c oxidase subunit 1, DNA barcoding, fish identification, Nayband National Park,

Persian Gulf

Received 23 September 2010; revision received 8 December 2010; accepted 22 December 2010

Introduction

FishBase reports 777 fish species from the Persian Gulf

(http://www.fishbase.org). These species have been

identified primarily on the basis of morphology. In mor-

phology-based taxonomy, the boundary between intra-

specific morphological variation and interspecific

morphological similarities may be blurred. Additional

drawbacks of morphology-based taxonomy include inap-

plicability of identification keys to all life stages, lack of

universal standard characters across taxa and require-

ment of high levels of expertise among a dwindling pool

of taxonomists. DNA barcoding was introduced in an

attempt to overcome some of these shortcomings (Hebert

et al. 2003). It aims to allow accurate and relatively simple

species identification based on the nucleotide sequence of

usually one short DNA fragment. A 650-bp fragment

from the 5¢ region of the mitochondrial COI gene has

been widely used for species-level identification across

a wide range of both invertebrate (Costa et al. 2007;

Mikkelsen et al. 2007) and vertebrate (Hebert et al. 2004;

Hajibabaei et al. 2006) animal species. Essentially, the

evolutionary rate of sequence change in this region of

the gene is such that its sequence in the genome of allCorrespondence: Elahe Elahi, Fax: +982166405141;

E-mails: [email protected], [email protected]

� 2011 Blackwell Publishing Ltd

Molecular Ecology Resources (2011) 11, 461–472 doi: 10.1111/j.1755-0998.2011.02989.x

members of a species lies within a cluster of very similar

sequences. The corresponding sequence in members of

other species, even in members of sister species, lies

outside that cluster.

Occasionally, the COI sequence data need to be aug-

mented by short sequences of other markers of mitochon-

drial or nuclear origin (Monaghan et al. 2006; Sonnenberg

et al. 2007). These findings allow the barcoding tool to be

used not only for species identification, but also for spe-

cies discovery. Any effective strategy for reducing cur-

rent rates of extinction would require characterization of

extant biodiversity. Conservative estimate for number of

eukaryotic species is 8–14 million (http://www.iucn.

org), whereas only approximately 2 million have been

described (Bisby et al. 2010). The notable fraction of spe-

cies that remain to be identified after more than two cen-

turies of taxonomic work attests both to the difficulty of

the task and to the need for implementation of methods

of higher throughput (Waugh 2007). In addition to sys-

tematics, the potentials of such protocols can be applied

to important challenges in ecology-, conservation- and

economy-related issues (Armstrong & Ball 2005; Mark-

mann & Tautz 2005; Savolainen et al. 2005; Smith et al.

2005; Barber & Boyce 2006; Asensio Gil 2007; Kyle & Wil-

son 2007; Ardura et al. 2010).

With regard to fish, approximately 15 000 morpho-

logically delineated marine species have been described

(Smith et al. 2003). Fish species are often difficult to

identify on the basis of morphology largely because of

dramatic phenotypic changes during development. For

this reason, they represent an appropriate taxon to be

targeted by the barcoding approach. So far, several

large-scale and small-scale barcoding projects have been

conducted on both marine and fresh water fishes (Ward

et al. 2005; Hubert et al. 2008; Steinke et al. 2009b; Val-

dez-Moreno et al. 2009). Fish larvae have also been iden-

tified on the basis of molecular markers (Pegg et al.

2006; Richardson et al. 2006). As of July 15, 2010, more

than 7700 fish species have been designated COI bar-

codes (Fish Barcode of Life website, http://www.fishbol.

org). Based on the results of some of the large-scale

studies, over 90% of the hundreds of morphologically

defined species in the cohorts could be distinguished by

their COI sequences (Ward et al. 2009). Additionally,

potential cryptic species and species incorrectly consid-

ered as separate species have been flagged. Many of the

new taxa were later confirmed by integrative taxonomic

reanalysis (Ward et al. 2009). Notably, data from a recent

study on 20 broadly distributed and economically

important inshore marine taxa suggested that nearly

half in fact represented sister species (Zemlak et al.

2009). These findings evidence the promise and poten-

tials of the barcoding approach for fish. Issues such as

hybridization, recent radiations, regional differentiation

in barcode sequences and nuclear copies of the barcode

region may reduce the reliability of barcoding for spe-

cies identification. However, it appears that these issues

are not severe for the majority of fish taxa (Ward et al.

2009).

In this study, we provide COI barcode sequences of

fishes residing at the Nayband National Park (NNP), Per-

sian Gulf, Iran. The park is located in the centre of the

northern coast of the Persian Gulf bounded by the lati-

tudes of 27�20¢N and 27�30¢N and the longitudes of

52�30¢E and 52�40¢E (Fig. 1). Nayband Bay includes sev-

eral ecosystems, such as mangrove forests and coral

reefs, and there is high biodiversity in the region. It was

classified a protected area in 1978 and promoted to a

National Park in 2004. Identification and conservation of

fish species of Nayband Bay is of ecological and eco-

nomic importance. Fishing is the major occupation of the

locals. Furthermore, the park is situated in proximity to

the South Pars Special Energy Zone (the Asaluyeh oil

refinery and petrochemical complexes) that is currently

discharging industrial wastes into the Bay. The ecological

significance of the Park, potential effects of oil industry-

related activities and the broad goals of The Fish Barcode

of Life campaign (http://www.fishbol.org) and The Con-

sortium for the Barcode of Life (http://www.barcoding.

si.edu) make it important that species residing in the

Nayband Bay be identified. Furthermore, this effort is a

step towards the larger goal of identifying all fish species

in the Persian Gulf. To the best of our knowledge, this is

the first report of barcoding data on fish species of the

Persian Gulf. Finally, we compared our data with meta-

data from other geographical regions such as Australia,

South Africa, India and China. This type of analysis may

flag hidden diversity because of geographic isolation.

Fig. 1 Position of Nayband Bay in the Persian Gulf. The filled

circle shows Haleh, an important fishing site in the bay, and the

filled rectangle shows the location of the Asaluyeh industrial

complex.


462 D N A B A R C O D I N G

Material and methods

Fish specimens

The fish were procured from Nayband Bay by web fish-

ing during July and December 2009 with the assistance of

local fishermen, mostly at the Haleh site (Fig. 1). Details

on collection dates and coordinates are available within

the project file ‘Nayband National Park Fishes’ (NNPF)

in BOLD (http://www.boldsystems.org) (Ratnasingham

& Hebert 2007). High quality photographs of the fish

were taken within hours of retrieval from water. All of

the specimens except one big grouper (from which tissue

sample and photos are available) are preserved as refer-

ence vouchers at the voucher maintenance center of Nat-

ural History Museum and Genetic Resources Bureau,

Department of Environment of IR Iran. Accession num-

bers are provided in BOLD. We aimed to sample at least

five adults of each species. Preliminary species identifica-

tions were carried out using FAO Area 51 identification

keys (Fischer & Bianchi 1984) and Smith’s Sea Fishes

(Smith et al. 2003). In cases where identifications were

ambiguous or particularly difficult, experienced taxono-

mists familiar with the local fauna were consulted. Scien-

tific names follow FishBase. Assignment of local names

was carried out by the fishermen.

Barcode sequencing

For most animals, a piece (5–20 mm3) of muscle was

removed from the dorsal side of each fish after photogra-

phy using flame-sterilized tools and placed in 99% etha-

nol for long-term preservation. The animals were then

placed on ice, frozen on site and transported as such to

location of voucher maintenance in Tehran. A 1-to 3-mm3

piece of each tissue sample was removed for DNA extrac-

tion, and the rest was kept as a tissue voucher. For a few

fish, tissue removal was performed in Tehran from frozen

animals. DNA extractions were performed using either

NucleoSpin96 (Machery-Nagel, Bethlehem, PA, USA) or

AccuPrep Genomic DNA Extraction kit (Bioneer, Daej-

eon, Korea) according to manufacturers’ instructions. A

652-bp segment of the COI gene was amplified by PCR

using FishR2_t1, FishF2_t1 or a primer cocktail contain-

ing FishR2_t1, FishF2_t1, FR1d_t1 and VF2_t1 primers

(Ivanova et al. 2007) under the following thermal profile:

94 �C for 10 min; five cycles of 94 �C for 45 s, 45 �C for

45 s, 72 �C for 45 s; 40 cycles of 94 �C for 45 s, 51 �C for

45 s, 72 �C for 45 s; 72 �C for 10 min. Each reaction mix-

ture contained 25 lL of 10% trehalose, 5 lL of 10· PCR

buffer, 2 lL of 50 mM MgCl2, 1 lL of 10 mM dNTP mix,

0.5 lL of each primer (10 pmol ⁄ lL), 1–2 lL of template

DNA and 1–2 U SmarTaq DNA polymerase (Cinnagen,

Tehran, Iran). Deionized water was added to obtain a

reaction volume of 50 lL. In some experiments, Platinum

Taq DNA polymerase (Invitrogen, Carlsbad, CA, USA)

was used, in which case extension and hold temperatures

were reduced to 68 �C. PCR products were visualized on

1% agarose gels. Sequencing was performed on samples

that produced a single band. A band of approximately

300 bp in addition to the band of expected size was

observed in a few cases. In such cases, DNA in the 650-bp

band was eluted using a gel extraction kit (Fermentas

International Inc., ON, Canada) and sequenced. Sequenc-

ing was performed bidirectionally using the ABI Big Dye

terminator chemistry and an ABI Prism 3700 instrument

(Applied Biosystems, Foster City, CA, USA).

To further query existence of cryptic species flagged

by COI sequences, 16S rDNA fragments of eight Gerres

filamentosus and two Plectorhinchus schotaf, as well as

three other specimens of the Gerres genus, were also

amplified and sequenced as described earlier. For

amplification of the 573-bp fragment encoding 16S

rRNA, the 16Sar-5¢ and 16Sbr-3¢ primers were used

(Ivanova et al. 2007).

Barcode sequence analysis

Chromatograms were initially inspected visually for

reading errors using the Sequencher software (Gene

Codes Corporation, Ann Arbor, MI, USA). The presence

of an extended open reading frame in COI sequences was

ascertained using ChromasPro version 1.41 (Technely-

sium, Australia). Sequences were then aligned in MEGA

4.0 (Tamura et al. 2007). The Fasta alignment was

uploaded onto the BOLD website. All sequence have

been deposited in GenBank; accession numbers for the

barcodes, sequences, chromatograms and primer

sequences are available at the project file ‘Nayband

National Park Fishes’ in BOLD. Calculation of Kimura

2-parameter (K2P) distances (Kimura 1980) and generation

of Neighbour-joining (NJ) tree (Saitou & Nei 1987) of K2P

distances were performed using the BOLD Management

and Analysis System. Calculation of K2P distances and

generation of NJ tree of 16S rDNA sequences were

performed using MEGA 4.0.

Meta-data analysis was performed using BOLD

Management and Analysis System to compare COI

sequences of selected species from this study with

conspecifics from Australia, India, China and South

Africa. Sequence information from the following pro-

jects in BOLD was used: ‘Fishes of Australia Container

Part I’, ‘DNA Barcoding the Indian Marine Fishes’,

‘Fishes From South China Sea’ and ‘Overlooked Fishes

in Marine Settings’. These projects were chosen on the

basis of inclusion of a reasonable number of species

that were also identified in our study. Three types of

divergence values were calculated. A commonly used


D N A B A R C O D I N G 463

global divergence value was calculated for each spe-

cies which was the average of all pairwise distances of

sequences belonging to the same species regardless of

location of origin. Regional divergences were calcu-

lated by averaging the distances of all conspecific

sequences from the same location. Finally, interregion-

al distances were calculated by averaging all distance

values obtained from comparing each of the sequences

from one location with all conspecific sequences of a

second location. A number of tentative molecular

operational taxonomic units (MOTUs) based on pat-

terns of sequence clustering and degrees of divergence

were inferred for each nominal species (Blaxter et al.

2005). A NJ tree inclusive of all the sequences was

also constructed. Finally, degrees of observed diver-

gence at different taxonomic levels of samples of this

study were compared with data from seven other

BOLD projects encompassing broader geographical

ranges. These projects included ‘Fishes of Australia

Container Part I’, ‘DNA Barcoding the Indian Marine

Fishes’, ‘Fishes From South China Sea’, ‘Aquarium

imports’, ‘Marine Fish of Mexico I’, ‘Marine Fish of

Mexico II’ and ‘Fishes of Pacific Canada Part I’.

Results

Taxonomic coverage

The specimens examined in this study included 187 indi-

viduals representing 76 species, 56 genera, 32 families, 11

orders and 2 classes of fish. The number of specimens per

species ranged from 1 to 8, with an average of 2.5. We

were unable to present exact taxonomic identification

for 22 individuals based on morphological features

(Table S1, Supporting information). Although assign-

ment of definitive genus and ⁄ or species nomenclature to

these individuals was not possible, they were assessed

based on morphology to likely belong to nine distinct

species. Table S2 (Supporting information) provides

detailed taxonomic information plus accession numbers

to BOLD and GenBank sequences for each of the 187

specimens.

Sequence analysis

COI barcode region sequences were retrieved from all

but three specimens that were identified by morphologi-

cal characters as Alectis indicus, Chirocentrus dorab and

Nemipterus bipunctatus. Reliable read lengths averaged

629 bp, and there were only four sequences shorter than

500 bps. No insertion ⁄ deletions, heterozygous sites or

stop codons were observed.

Mean intraspecific K2P divergence in the Nayband

samples was initially calculated to be 1.15% (range 0–

13.88%) while mean congeneric species K2P divergence

was 12.00% (range 6.19–20.23%) (Table 1). Significant cor-

relation between number of individuals per species and

maximum intraspecific variation was not observed

(regression line equation: y = 0.59x)1.1914, R2 = 0.1805).

The mean intraspecific distance observed was notably

higher than intraspecific distances reported for marine

(0.25–0.39%) (Ward et al. 2005; Steinke et al. 2009b) and

fresh water species (0.3–0.45%) (Hubert et al. 2008; Val-

dez-Moreno et al. 2009). Closer observation of the data

associated with the two species, Gerres filamentosus and

Plectorhinchus schotaf, which showed maximum intraspe-

cific divergences of 13.88% and 5.12%, respectively,

revealed that the specimens of each in fact formed two

clusters. The eight specimens of G. filamentosus formed

two clusters with intracluster K2P distances of 0% and

0.12%, while the mean intercluster distance was 13.84%

(standard deviation: 0.08). Diverged as they are, mem-

bers of the two clusters were closer to each other than to

members of any other species in our sample set. The two

P. schotaf specimens diverged by 5.12%. The clustering

pattern of 16S rDNA sequences of the G. filamentosus

samples was comparable to the pattern observed for their

COI sequences. Average intracluster sequence diver-

gences were 0.08% and 0.27%, while their mean interclus-

ter divergence was 4.24% which is more than 15 times

higher than the intracluster values. Divergence values for

comparisons between members of morphologically dis-

tinct Gerres species (hence, excluding comparisons

between any two G. filamentosus specimens) averaged

10.5% (range: 9.3–11.7%). A standard 16S divergence

Table 1 Kimura 2-parameter distances

between Nayband National Park

specimens at different taxonomic levels*

Number of

comparisons

Min.

distance (%)

Mean

distance (%)

Max.

distance (%)

SE

distance (%)

Within species 185 0 0.18 1.66 0.02

Within genus 76 6.19 12.00 20.23 0.45

Within family 888 10.88 17.43 24.56 0.08

Within order 9274 14.57 21.51 28.90 0.02

Within class 3439 16.20 22.77 34.41 0.04

*Specimens identified as Gerres filamentosus and Plectorhinchus schotaf based on morphol-

ogy were not included in the analysis.



value for species delineation to the best of our knowledge

does not exist. Divergences ranging between 8.46% and

21.12% have been reported for gadoid species (Di Finizio

et al. 2007). Divergences ranging from 4% (between

human and chimpanzee) to 41% (between human and

turkey) were reported for sequences of vertebrates (Kit-

ano et al. 2007). Expectedly, the two G. filamentosus clus-

ters show a minimal level of divergence characteristic of

closely related species. The rDNA sequences for the two

P. schotaf specimens differed only by presence of a single

nucleotide deletion in one of the sequences.

Removal of the COI sequences of the G. filamentosus

and P. schotaf specimens from the dataset caused reduc-

tion of mean intraspecific distance to 0.18% (range:

0–1.66%), a value that is closer to the results of previous

studies. The ratio between mean inter- and intraspecific

divergences was 10.4- and 66.7-fold, respectively, before

and after elimination of the two potential cryptic species.

After removal of the two species from the dataset, there

was no overlap between maximum K2P distance among

conspecific individuals (1.66%) and minimum distance

between individuals from congeneric species (6.19%). No

case of barcode sharing between two species was

observed. Table 1 presents genetic variation within the

NNP specimens through increasing taxonomic levels.

As already mentioned, the gap between mean intra-

specific and interspecific distances (0.18–12%) was much

more notable than the gap between mean interspecific

and intergeneric distances (12–17.43%). Lesser increases

of genetic variation through increasing taxonomic levels

were also observed. Above the species level, there were

in all cases some overlap between maximum K2P dis-

tance at one taxonomic level and minimum distance at

adjacent level.

Neighbour-joining analysis of COI sequences showed

that most species formed cohesive units with little

sequence variation (Fig. 2). The nearest neighbour to any

randomly chosen sequence was a sequence associated

with a conspecific specimen, unless the sequence was of

a species with only a single specimen sample. This is con-

sistent with correct taxonomic identification and suggests

that the COI barcode sequence of novel specimens can be

used for species identification. In this analysis, specimens

of G. filamentosus and P. schotaf produced deep internal

splits (arrows in Fig. 2) which are consistent with each

taxa being composed of two clusters as described earlier.

The G. filamentosus samples also produced two clusters in

the 16S rDNA NJ tree, with a clustering of sequences that

mirrored the COI tree (Fig. S1, Supporting information).

As in the COI tree, the rDNA sequences of these Gerres

specimens were still closer to each other than to other

Gerres sequences. The COI K2P-NJ tree can be accessed

using Data Analysis options in the NNPF project page on

the BOLD website.

Meta-data analysis on 16 species produced interesting

results confirming previous reports of existence of regio-

nal divergences of sequences that may be indicative of

overlooked fish species (Zemlak et al. 2009). The results

of the analysis are presented in Table 2. Evidently, regio-

nal divergences were generally low, and only those of

Argyrops spinifer from Australia, Rastrelliger kanagurta

from India and, Nemipterus japonicus from Iran exceeded

0.5%. The A. spinifer specimens from Australia formed

two distinguishable clusters in the NJ tree (Fig. 3 and

Fig. S2, Supporting information). This observation and

the fact that their intraregional divergence values showed

an unusually high standard deviation (1.68 ± 1.26%) are

consistent with the idea that they may represent two

lineages. Contrary to within-region comparisons,

between-region comparisons revealed a broad range of

divergence values, from a minimum of 0 (Megalaspis

cordyla from Iran and India) to a maximum of 16.05%

(Platycephalus indicus from China and South Africa).

Interregion distances varied greatly depending on both

the taxon being examined and the choice of locations.

P. indicus, Rhabdosargus sarba and Lethrinus nebulosus from

three or more geographic regions each showed high

degrees of allopatric sequence divergence in all compari-

sons, whereas Scombroides tol and Sphyraena putnamae

seemed to comprise almost uniform populations

throughout the three or more regions in which they were

reported. Sympatric divergences of some species slightly

exceed allopatric divergences; examples include N. japo-

nicus (Ir > Ir ⁄ In) and A. spinifer (Ir > Ir ⁄ SA, SA > Ir ⁄ SA).

Although the sample sizes are small, the small differ-

ences may indicate that individuals of these species from

distant regions may constitute populations intercon-

nected by frequent gene flow or populations recently

founded from the same source population and gene pool.

Divergences between specimens from Iran and speci-

mens from India were generally low (0–0.51%), which is

probably explained by the fact that the Persian Gulf is

connected to the open waters only through the western

parts of the Indian Ocean. Generally, there was no simple

correlation between geographical distance and sequence

divergence of species. Sequences from Iran and South

Africa showed the highest similarity when considering

Scomberomorus commerson or A. spinifer, but were most

distantly diverged when R. sarba or Chanos chanos were

examined. Irregularity of divergence patterns is well

illustrated in Fig. 4. It can also be inferred from the NJ

tree of the specimen sequences (Fig. S2, Supporting infor-

mation). A 3.5% divergence threshold on global diver-

gence values (Zemlak et al. 2009) suggested that three of

the 16 species analysed likely represented more than one

MOTU. Most importantly, examination of interregional

divergences using the same threshold revealed that an

additional three species also represented more than one



MOTU. The six species associated with multiple MOTUs

based on interregional analysis together represented

fourteen MOTUs; P. indicus alone constituted four

MOTUs. Standard deviation values were generally low

for all interregional comparisons, irrespective of the

mean value. This suggests that the division of conspecific

sequences into one or more MOTUs based on these mean

distances is robust. On the other hand, high values of

average global divergence were always accompanied by

correspondingly high SD values. This, along with low

within-region distances and low SD for both within-

region and between-region comparisons, strongly

Fig. 2 Neighbour-joining tree for specimens in the Nayband National Park Fishes project. Arrows emphasize intraspecific splits.



implies that sequence divergences are geographically

structured.

Table 3 compares mean divergence values at different

taxonomic levels from eight barcoding projects on marine

fishes, including the NNPF project. Although sample sizes

varied notably in the different projects (138–1638),

the values at each taxonomic level across the eight

projects were remarkably similar. The somewhat lower

Fig. 2 Continued



Tab

le2

Intr

areg

ion

al,i

nte

rreg

ion

alan

dg

lob

ald

ista

nce

bet

wee

nm

emb

ers

of

16se

lect

edsp

ecie

s*

Sp

ecie

s

Glo

bal

div

erg

ence

(mea

n±

SD

)

Intr

areg

ion

ald

iver

gen

ces

(mea

n±

SD

)In

terr

egio

nal

div

erg

ence

s(m

ean

±S

D)

No

.

MO

TU

s†Ir

InC

hA

uS

AIr

⁄In

Ir⁄C

hIr

⁄Au

Ir⁄S

AIn

⁄Au

Ch

⁄Au

Ch

⁄SA

Au

⁄SA

Arg

yrop

s

spin

ifer

2.99

±2.

66(1

4)0.

16±

0.17

(6)

1.68

±1.

26(5

)0.

20±

0.09

(3)

5.66

±0.

780.

13±

0.11

5.33

±0.

282:

Ir,S

A|

Au

Cha

nos

chan

os0.

36±

0.27

(6)

(1)

0±

0(3

‡)0

(2)

0.31

±0

0.8

±0.

030.

48±

0.02

1

Chi

roce

ntr

us

dora

b

5.05

±5.

42(4

)0.

15(2

)(1

)(1

)0.

08±

0.11

10.0

2±

0.13

9.96

2:Ir

,Ch

|A

u

Dec

apte

rus

russ

elli

0.10

±0.

09(6

)(1

)0.

06±

0.08

(5)

0.18

±0.

071

Epi

nep

helu

s

chlo

rost

igm

a

0.10

±0.

09(3

)(1

)0.

15(2

)0.

08±

0.11

1

Let

hrin

us

neb

ulo

sus

3.12

±2.

33(1

1)0.

27±

0.25

(3)

0.32

±0.

26(5

)0.

11±

0.10

(3)

2.21

±0.

175.

29±

0.66

5.75

±0.

202:

Ir,A

u|

SA

Meg

alas

pis

cord

yla

0±

0(9

)0

±0

(4)

0±

0(5

)0

±0

1

Nem

ipte

rus

japo

nic

us

0.39

±0.

28(6

)0.

59(2

)0.

31±

0.24

(4)

0.42

±0.

331

Pla

tyce

phal

us

indi

cus

9.46

±6.

06(1

8)0.

32±

0.18

(4)

0±

0(7

)0.

18±

0.16

(5)

0(2

)15

.78

±0.

1310

.61

±0.

144.

05±

0.12

12.0

5±

0.00

16.0

5±

010

.95

±0

4:Ir

|C

h|

Au

|S

A

Pse

ttod

es

eru

mei

0.88

±1.

12(6

)(1

)0.

15±

0.10

(4)

(1)

0.08

±0.

092.

352.

43±

0.09

1:Ir

,In

,Au

Ras

trel

lige

r

kan

agu

rta

0.46

±0.

28(8

)0.

3±

0.17

(5)

0.72

±0.

18(3

)0.

51±

0.31

1

Rha

bdos

argu

s

sarb

a

4.66

±4.

50(1

7)0

±0

(6)

0.10

±0.

09(3

)0.

12±

0.10

(5)

0±

0(3

)2.

91±

0.09

2.57

±0.

1311

.89

±0.

062.

31±

0.11

11.5

8±

0.12

10.0

1±

0.10

2:Ir

,Ch

,

Au

|S

A

Sau

rida

tum

bil

0.10

±0.

10(9

)0.

13±

0.10

(5)

0.08

±0.

08(4

)0.

10±

0.11

1

Sco

mbe

roid

esto

l0.

14±

0.13

(10)

0.15

(2)

0.06

±0.

08(5

)0.

15(2

)(1

)0.

11±

0.10

0.15

±0.

130.

14±

0.20

0.11

±0.

100.

34±

0.12

0.28

±0

1

Sco

mbe

rom

oru

s

com

mer

son

2.03

±1.

58(9

)(1

)(1

)0.

50±

0.34

(5)

0(2

)3.

643.

55±

0.22

0.31

±0

0.65

±0.

423.

64±

03.

55±

0.20

2:Ir

,SA

|

Ch

,Au

Sph

yrae

na

putn

amae

0.07

±0.

08(9

)(1

)0.

04±

0.07

(7)

(1)

0.17

±0.

060.

150.

02±

0.06

1

SD

,sta

nd

ard

dev

iati

on

;Ir,

Iran

;In

,In

dia

;Ch

,Ch

ina;

Au

,Au

stra

lia;

SA

,So

uth

Afr

ica;

MO

TU

,mo

lecu

lar

op

erat

ion

alta

xo

no

mic

un

it.

*Nu

mb

ers

inp

aren

thes

esar

en

um

ber

of

ind

ivid

ual

so

fre

lev

ant

spec

ies.

†Nu

mb

ero

fM

OT

Us

isb

ased

on

ath

resh

old

dis

tan

cev

alu

eo

f3.

5%.F

or

spec

ies

wit

h‡1

MO

TU

,reg

ion

ald

istr

ibu

tio

no

fM

OT

Us

con

sist

ent

wit

hth

ed

ata

isp

rese

nte

d.

‡Tw

ofr

om

Au

stra

lia,

on

efr

om

Tai

wan

.



intraspecific variation of the NNP specimens is probably

due to the fact that Nayband Bay is a relatively small

area. Smaller increase of variation at higher taxonomic

levels observable in all the projects is likely affected by

the nature of the K2P distance-based model. For example,

this model does not consider back or recurrent muta-

tions, the frequency of which increases with longer aver-

age times of separation between individuals within

Fig. 4 Representation of intraregional,

interregional and global divergences for

16 selected fish species from five regions

of Indo-West Pacific Ocean. The species

are indicated, and the various divergences

are colour coded. Notice the low within-

region distances and the irregular pattern

of interregion divergences, which in some

cases are higher than the respective global

divergence values. For example, compare

the bars relating to Iran ⁄ China and Iran ⁄South Africa distances for Rhabdosargus

sarba and Platycephalus indicus.

Fig. 3 Neighbour-joining tree relating to

Argyrops spinifer specimens. Splits reflect-

ing sympatric divergence among speci-

mens from Australia and allopatric

divergence among specimens of three

geographic regions are evident.

Table 3 Mean Kimura 2-parameter distances at different taxonomic levels from eight barcoding projects on marine fishes

NNPF AUSA WLIND FSCS MM TZFPC TZAIC

Within species 0.17 0.45 0.22 0.22 0.36 0.23 0.47

Within genus 12.03 7.76 15.42 16.05 12.87 3.70 12.08

Within family 17.41 16.90 17.35 20.72 18.85 13.92 20.24

Within order 21.51 21.98 23.70 24.37 23.32 22.06 23.74

Within class 22.78 24.64 24.53 25.32 25.29 24.70 23.96

Sample size* 171 1454 138 711 610 1225 1638

NNPF, Nayband National Park Fishes (Gerres filamentosus and Plectorhinchus schotaf data removed); AUSA, Fishes of Australia Container

Part I; WLIND, DNA Barcoding the Indian Marine Fishes; FSCS, Fishes From South China Sea; MM, Marine Fish of Mexico I and Marine

Fish of Mexico II, merged; TZFPC, Fishes of Pacific Canada Part I; TZAIC, Aquarium imports.

*Sample size indicates the number of sequences used in the analysis of the respective project. The BOLD default option included only

the sequences longer than 420 bp.



higher taxonomic levels when compared with indivi-

duals within lower levels.

Discussion

Specimens examined in this study belonged to 76 species

from 32 families. We are in communication with expert

ichthyologists to identify the 22 specimens we were till

now unable to identify at the species level. Researchers of

a project performed in NNP in 2007 (‘Assessment of the

biological resources of the Nayband bay’; Department of

Environment of IR Iran, internal report) reported exis-

tence of 100–150 fish species in the bay. Based on these

results, the sample set studied here may represent 1 ⁄ 2 to

2 ⁄ 3 of the bay’s fish biodiversity. More extensive sam-

pling in different seasons and on site identification to

facilitate distinction between similar species is suggested

for further work in this region. Three of the species stud-

ied here, Cephalopholis hemistiktos, Epinephelus bleekeri and

Epinephelus coioides are given ‘Near Threatened’ status,

and Chaenogaleus macrostoma is classified ‘Vulnerable’ by

the International Union for Conservation of Nature

(IUCN).

All of the 76 species examined in this study can be

distinguished using COI barcodes. Two species, Gerres

filamentosus and Plectorhinchus schotaf, showed high

mean sympatric divergences of 7.46% and 5.12%,

respectively, suggesting each may include cryptic spe-

cies. 16S rDNA data were consistent with the proposal

that the NNP G. filamentosus samples represented two

closely related but distinct species. 16S rDNA

sequences evolve more slowly than COI sequences

(Hebert et al. 2003), and the data on these sequences

were not revealing with regard to the possible split

between the two P. schotaf specimens. Final assessment

of G. filamentosus and P. schotaf status awaits sequence

results of more genes from more specimens and

detailed examination of other categories of characters

including morphological ones.

After removal of G. filamentosus and P. schotaf from

the dataset, the gap between maximum intraspecific

divergence (1.66%) and minimum divergence between

congeneric taxa (6.19%) is notable, thus corroborating

that species identification of new specimens can be made

on the basis of COI barcode sequences. It has been recom-

mended that a sequence divergence that is ten times or

more than the average of observed within species varia-

tions can be used as a threshold to flag cryptic species

(Hebert et al. 2004). The average within species variation

of 0.18% in our study suggests that distances ‡1.8% may

be appropriate for flagging cryptic fish species of NNP.

In fact, maximum observed intraspecific divergence

(1.66%) was below this threshold, and minimum

interspecific divergence (6.19%) was above it. Notably,

application of the recommended criterion would sug-

gest that G. filamentosus and P. schotaf specimens both

contain cryptic species, consisting with our analysis

above.

NJ trees are usually not considered valid tools for

inference of phylogeny. Furthermore, it is widely agreed

that no single gene, including the mitochondrial COI

gene, provides adequate information for reconstructing

evolutionary relationships (Ward et al. 2005). Nonethe-

less, the NJ tree built here based on COI sequences clearly

harbours some strong phylogenetic signals. All conspe-

cific sequences and almost all congeneric and confamilial

taxa clustered together. Expectedly, the three elasmo-

branch species form a distinct cluster separate from all

teleost fishes.

Comparison of sequences of 16 selected species from

this study with conspecifics from one to four other pro-

jects on BOLD revealed high degrees of interregion diver-

gence within some taxa. Using the stringent threshold

value of 3.5%, six species were flagged as candidates for

representing cryptic species. Pairwise interregional com-

parisons were more informative than global divergence

values for flagging potentially cryptic species. Data on

Psettodes erumei illustrate an important methodological

point. Global divergence for this species was 0.88%. The

interregion distance data show that the Australian line-

age diverges from each the Iranian and Indian lineages

by more than 2%, whereas the Iranian and Indian lin-

eages diverge by only 0.08%. Because divergences with

the Australian lineage did not reach 3.5%, we proposed

only one MOTU for Psettodes erumei. However, use of a

universal threshold may not be appropriate. In fact, some

studies have applied a 2% value to screen cryptic species

(Steinke et al. 2009a). Whatever the appropriate thresh-

old for species delineation of Psettodes erumei specimens

from different regions may be, the averaging process

used to calculate their global divergence masked the high

variation imbedded in the Australian lineage because 10

out of the 15 sequence comparisons were Iranian ⁄ Indian

P. erumei comparisons. This effect is always possible

when collections of different sizes from different regions

are used for global comparisons. Cryptic species in multi-

regional studies are more likely to be detected by pair-

wise interregional comparisons. When collection data are

not available as may be the case for old specimens, stan-

dard deviation values associated with mean divergences

may signal unapparent variations.

Finally, we hope that this barcode study on Persian

Gulf fishes will quickly be followed by others. We hope

that accumulation of biodiversity data will help planning

and monitoring of conservation programmes in the

region. The promise of DNA barcoding for coral reef con-

servation has already been remarked upon (Neigel et al.

2007). Intensity of industrial and commercial activities in



the Persian Gulf, specially the region surrounding the

Nayband National Park, makes the biodiversity of the

region vulnerable to gradual and accidental degradation

and calls for prompt cataloguing of extant species.

Results of barcoding projects are expected to be valuable

for planning conservation and rehabilitation schemes for

the endangered ecosystems. Owing to high species rich-

ness and the legal status of the region as a National Park,

Department of Environment of IR Iran has been

requested to consider it a candidate for a comprehensive

location-centred case study of DNA barcoding at the

level of whole ecosystems.

Acknowledgements

We thank the local fishermen of NNP for assisting in sample col-

lection, the guards of NNP for providing accommodations, and

Dr Yasami, Dr Moore, Dr Krupp and Mr Dehghani for assistance

in fish identification. We also acknowledge and thank Dr Robert

Hanner, Ms Heather Braid and Dr Nasser Ghaemi for their assis-

tance. Finally, we acknowledge the Department of Environment

of IR Iran and PhD student support for Hosseinali Asgharian

from the College of Science, University of Tehran, for funding

this research.

References

Ardura A, Linde AR, Moreira JC, Garcia-Vazquez E (2010) DNA barcod-

ing for conservation and management of Amazonian commercial fish.

Biological Conservation, 143, 1438–1443.

Armstrong KF, Ball SL (2005) DNA barcodes for biosecurity: invasive spe-

cies identification. Philosophical Transactions of the Royal Society of Lon-

don, Series B: Biological Sciences, 360, 1813–1823.

Asensio Gil L (2007) PCR-based methods for fish and fishery

products authentication. Trends in Food Science and Technology, 18, 558–

566.

Barber P, Boyce SL (2006) Estimating diversity of Indo-Pacific coral reef

stomatopods through DNA barcoding of stomatopod larvae. Proceed-

ings of the Royal Society of London, Series B: Biological Sciences, 273, 2053–

2061.

Bisby FA, Roskov YR, Orrell TM et al. (eds) (2010) Species 2000 & ITIS

Catalogue of Life: 2010 Annual Checklist, 10 (Species 2000: Reading,

UK). Available at: http://www.catalogueoflife.org/annual-checklist/

2010.

Blaxter M, Mann J, Chapman T et al. (2005) Defining operational

taxonomic units using DNA barcode data. Philosophical Transactions

of the Royal Society of London, Series B: Biological Sciences, 360, 1935–

1943.

Costa FO, deWaard JR, Boutillier J et al. (2007) Biological identification

through DNA barcodes: the case of Crustacea. Canadian Journal of Fish-

eries and Aquatic Sciences, 64, 272–295.

Di Finizio A, Guerriero G, Russo GL, Ciarcia G (2007) Identification of

gadoid species (Pisces, Gadidae) by sequencing and PCR–RFLP analy-

sis of mitochondrial 12S and 16S rRNA gene fragments. European Food

Research and Technology, 225, 337–344.

Fischer W & Bianchi G (eds) (1984) FAO Species Identification Sheets for

Fishery Purposes. Western Indian Ocean (Fishing Area 51). Food and Agri-

cultural Organization of the United Nations, Rome.

Hajibabaei M, Singer GA, Hickey DA (2006) Benchmarking DNA bar-

codes: an assessment using available primate sequences. Genome, 49,

851–854.

Hebert PD, Cywinska A, Ball SL, deWaard JR (2003) Biological identifica-

tions through DNA barcodes. Proceedings of the Royal Society of London,

Series B: Biological Sciences, 270, 313–321.

Hebert PD, Stoeckle MY, Zemlak TS, Francis CM (2004) Identification of

birds through DNA barcodes. PLoS Biology, 2, e312.

Hubert N, Hanner R, Hom E et al. (2008) Identifying Canadian freshwater

fishes through DNA barcodes. PLoS One, 3, e2490.

Ivanova NV, Zemlak TS, Hanner RH, Hebrt PD (2007) Universal

primer cocktails for fish DNA barcoding. Molecular Ecology Notes, 7,

544–548.

Kimura M (1980) A simple method for estimating evolutionary rate of

base substitutions through comparative studies of nucleotide

sequences. Journal of Molecular Evolution, 15, 111–120.

Kitano T, Umetsu K, Tian W, Osawa M (2007) Two universal primer sets

for species identification among vertebrates. International Journal of

Legal Medicine, 121, 423–427.

Kyle C, Wilson C (2007) Mitochondrial DNA identification of game and

harvested freshwater fish species. Forensic Science International, 166, 68–

76.

Markmann M, Tautz D (2005) Reverse taxonomy: an approach towards

deter mining the diversity of meiobenthic or ganisms based on ribo-

somal RNA signature sequences. Philosophical Transactions of the Royal

Society of London, Series B: Biological Sciences, 360, 1917–1924.

Mikkelsen NT, Schander C, Willassen E (2007) Local scale DNA bar-

coding of bivalves (Mollusca): a case study. Zoologica Scripta, 36, 455–

463.

Monaghan MT, Balke M, Pons J, Volger AP (2006) Beyond barcodes: com-

plex DNA taxonomy of a South Pacific Island radiation. Proceedings of

the Royal Society of London, Series B: Biological Sciences, 273, 887–893.

Neigel J, Domingo A, Stake J (2007) DNA barcoding as a tool for coral reef

conservation. Coral Reefs, 26, 487–499.

Pegg GG, Sinclair B, Briskey L, Aspden WJ (2006) MtDNA barcode identi-

fication of fish larvae in the southern Great Barrier Reef, Australia.

Scientia Marina, 70(Suppl. 2), 7–12.

Ratnasingham S, Hebert PD (2007) BOLD: the Barcode of Life Data

System (http://www.barcodinglife.org). Molecular Ecology Notes, 7,

355–364.

Richardson DE, Vanwye JD, Exum AM, Cowen RK, Crawford DL (2006)

High-throughput species identification: from DNA isolation to bioin-

formatics. Molecular Ecology Notes, 7, 199–207.

Saitou N, Nei M (1987) The neighbour-joining method: a new method for

reconstructing evolutionary trees. Molecular Biology and Evolution, 4,

406–425.

Savolainen V, Cowan RS, Volger AP, Roderick GK, Lane R (2005)

Towards writing the encyclopaedia of life: an introduction to DNA bar-

coding. Philosophical Transactions of the Royal Society of London, Series B:

Biological Sciences, 360, 1805–1811.

Smith JL, Smith MM, Heemstra PC (2003) Smith’s Sea Fishes. Struik, Cape

Town.

Smith MA, Fisher BL, Hebert PD (2005) DNA barcoding for effective bio-

diversity assessment of a hyperdiverse ar thropod group: the ants of

Madagascar’. Philosophical Transactions of the Royal Society of London,

Series B: Biological Sciences, 360, 1825–1834.

Sonnenberg R, Nolte AW, Tautz D (2007) An evaluation of LSU rDNA

D1-D2 sequences for their use in species identification. Frontiers in Zool-

ogy, 4, doi:10.1186/1742-9994-4-6.

Steinke D, Zemlak TS, Hebert PD (2009a) Barcoding Nemo: DNA-based

Identification for the Ornamental Fish Trade. PLoS One, 4, e6300.

Steinke D, Zemlak TS, Boutillier JA, Hebert PD (2009b) DNA barcoding of

Pacific Canada’s Fishes’. Marine Biology, 156, 2641–2647.

Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: Molecular Evolu-

tionary Genetics Analysis (MEGA) software version 4.0. Molecular Biol-

ogy and Evolution, 24, 1596–1599.

Valdez-Moreno M, Ivanova NV, Elias-Gutierrez M, Contreras-Balderas S,

Hebert PD (2009) Probing biodiversity in freshwater fishes from Mex-

ico and Guatemala with DNA barcodes. Journal of Fish Biology, 74, 377–

402.



Ward RD, Zemlak TS, Innes BH, Last PR, Hebert PD (2005) DNA barcod-

ing Australia’s fish species. Philosophical Transactions of the Royal Society

of London, Series B: Biological Sciences, 360, 1847–1857.

Ward RD, Hanner R, Hebert PD (2009) The campaigh to DNA barcode all

fishes, FISH-BOL. Journal of Fish Biology, 74, 329–356.

Waugh J (2007) DNA barcoding in animal species: progress, potential and

pitfalls. Bioessays, 29, 188–197.

Zemlak TS, Ward RD, Connell AD, Holmes BH, Hebert PD (2009) DNA

barcoding reveals overlooked marine fishes. Molecular Ecology

Resources, 9(Suppl. 1), 237–242.

Supporting Information

Additional supporting information may be found in the

online version of this article.

Fig. S1 NJ tree of 16S rDNA sequences for NNPF samples

of the genus Gerres. NNPF1135 and NNPF1137: G. oyena,

NNPF1176: Gerres sp., the other eight indivuals (the two

top clusters): G. filamentosus.

Fig. S2 NJ tree of COI sequences of 16 selected species

from NNPF and four other fish barcoding projects.

Table S1 NNPF samples with incomplete or ambiguous

taxonomic identification.

Table S2 Detailed taxonomic and specimen data of

NNPF samples.

Please note: Wiley-Blackwell are not responsible for the

content or functionality of any supporting information

supplied by the authors. Any queries (other than missing

material) should be directed to the corresponding author

for the article.



Cytochrome c oxidase subunit 1 barcode data of fish of the Nayband National Park in the Persian Gulf...

Documents

Transcript of Cytochrome c oxidase subunit 1 barcode data of fish of the Nayband National Park in the Persian Gulf...