Gobiids are hyperdiverse compared with other teleost groups, with about 2,000 species occurring in marine, freshwater, and blackish habitats, and they show a remarkable variety of morphologies and ecology. Testing the effectiveness of DNA barcodes on species that have emerged as a result of radiation remains a major challenge in evolutionary biology. Here, we used the cytochrome c oxidase subunit 1 (COI) sequences from 144 species of gobies and related species to evaluate the performance of distance-based DNA barcoding and to conduct a phylogenetic analysis. The average intra-genus genetic distance was considerably higher than that obtained in previous studies. Additionally, the interspecific divergence at higher taxonomic levels was not significantly different from that at the intragenus level, suggesting that congeneric gobies possess substantial interspecific sequence divergence in their COI gene. However, levels of intragenus divergence varied greatly among genera, and we do not provide sufficient evidence for using COI for cryptic species delimitation. Significantly more nucleotide changes were observed at the third codon position than that at the first and the second codons, revealing that extensive variation in COI reflects synonymous changes and little protein level variation. Despite clear signatures in several genera, the COI sequences did resolve genealogical relationships in the phylogenetic analysis well. Our results support the validity of COI barcoding for gobiid species identification, but the utilization of more gene regions will assist to offer a more robust gobiid species phylogeny.
INTRODUCTION
Gobies (family Gobiidae; suborder Gobioidei) are incomparable among vertebrates in their capacity to adapt and diversify, which has led to adaptive radiation and rapid speciation (Zander, 2011). Gobiid fishes are hyperdiverse compared with other teleost groups, with approximately 2,000 species in 210 genera occurring in marine, freshwater, and blackish habitats. These fish show remarkable morphological and ecological variety (Nelson, 2006; Zander, 2011). Gobiid fishes are globally distributed (Nelson, 2006) and frequently represent a dominant component of coral reefs and coastal fish communities throughout much of their range, accounting for >50% of the energy flow in some coral reef habitats (Herler et al., 2011). Despite their evolutionary and ecological importance, the phylogenetic relationships among species within Gobiidae and their location within Gobioidei are still poorly understood (Murdy, 1989; Parenti and Thomas, 1998; Thacker and Schaefer, 2000; Larson, 2001). To date, the classification of gobies still remains largely reliant on external morphology (Pezold, 1993; Akihito et al., 2000; Nelson, 2006), and diagnostic characters separating species are subtle and problematic.
Molecular biology has contributed to addressing taxon identification and phylogenetic relationship questions. Mitochondrial DNA (mtDNA) markers have historically formed the core of most molecular systematic analyses and are still the most widely used for reconstructing phylogeny (Brown et al., 1979; Moore, 1995; Johns and Avise, 1998); this is probably due to their single copy nature and relative ease of sequencing (Moore, 1995). Genetic divergence is also enhanced by the higher rate of sequence evolution in vertebrate mtDNA compared to that of nuclear coding regions (Johns and Avise, 1998). However, the choice of a suitable gene is crucial for identification and phylogenetic reconstruction among closely related species (Brown et al., 1979; Moore, 1995; Johns and Avise, 1998), because different parts of the mtDNA genome evolve at different rates (Avise and Ellis, 1986; Roques et al., 2006).
Mitochondrial cytochrome oxidase subunit I (COI) could serve as a rapid and reliable barcoding marker for identifying species and for discovering new species across the entire animal kingdom (Hebert et al., 2003). Although skepticism has frequently been expressed (Ebach and Holdrege, 2005; Will et al., 2005), DNA barcoding based on COI has been successful to identify species across a wide array of taxa over the last decade (e.g., Hebert et al., 2004; Clare et al., 2007; Hubert et al., 2008; Feng et al., 2011). A clear gap should exist between intra- and interspecific COI sequence divergence with about a 20-fold difference for DNA barcoding to be perfectly effective in delimitating species (Hebert et al., 2003). A standard sequence threshold can be projected to outline species boundaries by employing this barcoding gap. However, utilizing such a threshold value may be challenging, particularly when attempts do not include numerous specimens, such as for critically endangered taxa.
Only a few studies have addressed gobioid interrelationships based on molecular data (e.g., Akihito et al., 2000; Wang et al., 2001; Thacker, 2003, 2009; Thacker and Hardman, 2005). These studies used different taxon and nucleotide sampling methods. Yet, testing the effectiveness of COI DNA barcodes on species that have emerged as a result of radiation, such as gobies, remains a major challenge in evolutionary biology. Here, we sequenced the COI of 48 species collected from South Korea (
Table 1
) to evaluate the performance of distance-based DNA barcoding for phylogenetic analyses. We specifically aimed to provide novel data on a comparison of pairwise divergence levels among species in the same genus vs. species in different genera. GenBank sequences were also included in the analyses to use a dataset with large taxonomic coverage (n=144 species) (
Table 2
).
MATERIALS AND METHODS
- Sample collection
Fish were collected using seine and dip nets from January to November 2011 from 21 sites across freshwater systems, coastal areas, and the ocean near South Korea (
Table 1
). Specimens were identified based on morphological characters. Entire bodies of all individuals were preserved in 95% ethanol, and 44 nominal species were sequenced for COI gene fragments.
- DNA isolation, amplification, and sequencing
We used the Wizard Genomic DNA purification kit (Promega, Madison, WI, USA) to extract genomic DNA from the right pectoral fin of each fish specimen. The COI was amplified using gobiid-specific primers: GOBYF7558 (forward) 5′- TTT GCW ATT ATG GCW GGA TTT G-3′ and GOBYB 8197 (reverse) 5′-ATT ATT AGG GCG TGG TCG TGG-3′ (Thacker, 2003) and COI fish universal primers, FF2d (forward) 5′-TTC TCC ACC AAC CAC AAR GAY ATY GG- 3′ and FF1d (reverse) 5′-CAC CTC AGG GTG TCC GAA RAA YCA RAA-3′ (Ivanova et al., 2007). Each polymerase chain reaction (PCR) amplification was carried out in a 50 μL reaction volume composed of ~75 ng DNA extract, 0.25 mM of each deoxynucleotide, 0.25 mM of each forward and reverse primer, 3 mM MgCl2, 1×PCR buffer, and 0.25 units of Taq DNA polymerase (Solgent, Daejeon, Korea). GenePro (BIOER) was used to amplify the COI with the following program: 94℃ for 10 min, 35 cycles of 30 s at 94℃, 30 s at 54℃ (for GOBYF7558-GOBYB8197) and 52℃ (for FF2d- FF1d), 30 s at 72℃ and final elongation at 72℃ for 10 min. PCR products were loaded on 1% agarose gels containing 0.003% ethidium bromide and visualized using the GelDoc- It TM Imaging System (UVP). Amplifications were considered successful when a expected sized band was observed on the agarose gel. PCR products were cleaned using a PCR purification kit (Solgent). The COI was sequenced directly using the BigDye-Terminator V3.1 kit (Applied Biosystems, Foster City, CA, USA) and an ABI3730XL sequencer at Genotech (Daejeon, Korea).
- Sequence data analyses
Complementary DNA sequences were assembled using the Bioedit 5.0.9 sequence-editing software (Hall, 1999). Sequences were aligned using Clustal X 2.0 default settings (Larkin et al., 2007). Alignments were translated to amino acids under the vertebrate mitochondrial option using MEGA 5 (Tamura et al., 2011) to detect frameshift mutations and premature stop codons, which may indicate the presence of pseudogenes. The Genbank accession numbers of newly determined sequences were JX679021-JX679066 and are listed in Table 1. Genetic distances were calculated to quantify sequence divergences among species using both p distance and Kimura two-parameter (K2P) models (1,000 bootstrapping) (Kimura, 1980), as implemented in MEGA. Rates of synonymous and nonsynonymous substitutions were also calculated with MEGA using both standard and modified (at 1.4 standard errors with 1,000 bootstrapping samples) Nei- Gojobori models (Nei and Gojobori, 1986; Nei and Kumar, 2000). Genetic distances were calculated at intrageneric, intrasubfamilial, intrafamilial, and interfamilial levels. Altogether, 10,296 pairwise distances were compared in this study. The degree of sequence conservation per site,
R
seq
, was defined as
R
seq
=2-(Σ
p
log
2
p
) (Ward and Holmes, 2007),
Details of gobioid fish species analyzed; data comprised of species name, voucher number, locality, GPS coordinate, collection date, and GenBank accession number; F, E and O in bracket on each location indicates habitat information such as freshwater, estuary and ocean, respectively
Details of gobioid fish species analyzed; data comprised of species name, voucher number, locality, GPS coordinate, collection date, and GenBank accession number; F, E and O in bracket on each location indicates habitat information such as freshwater, estuary and ocean, respectively
List of reference species used in this study
List of reference species used in this study
Continued
where
p
is the observed frequency of each base at a particular position and the maximal degree of conservation was 2, which was achieved when all nucleotides at a particular site in the 144 species were the same.
A Bayesian inference (BI) tree was established using MrBayes 3.1.2 (Ronquist and Huelsenbeck, 2003) with two outgroup species from the Family Rhyacichthydae,
Protogobius attiti
and
Ryacichthys guilberti
. The best-fit model of DNA sequence evolution was chosen using ModelTest 3.8 (Posada and Crandall, 1998) and Akaike information criteria; the chosen model was GTR+I+G. The analysis was run for 10 million generations with sampling of one tree every 500 generations. Two independent Markov Chain Monte Carlo runs were conducted simultaneously. The first 1,000 trees of each run were discarded as burn-in.
RESULTS
The COI genes of each species were confidently aligned, and the equivocal bases at each end were trimmed to yield a final sequence of 542 bp. No indels were detected. Translation of the sequences did not reveal frame-shift mutations or premature stop codons, confirming that our amplified fragments were functional. Among the 542 nucleotide positions, 245 were polymorphic, and 230 were parsimony informative. The proportion of T, C, A, and G bases for all 144 sequences was 30.5%, 28.1%, 23.1%, and 18.2%, respectively. The GC content was relatively higher at the first codon base (56.3%) than that at the second (43.1%) or third (39.6%). The degree of conservation (
R
seq
) was calculated for each base of the 542 nucleotides; the most common and maximum value was 2, which was achieved when all nucleotides at a particular site
Estimated evolutionary parameters (×100) for the nucleotide substitutions in the cytochrome oxidase c subunit 1 (COI) barcoding region from 144 gobioid fish speciesK2P, Kimura-2-parameter.
Estimated evolutionary parameters (×100) for the nucleotide substitutions in the cytochrome oxidase c subunit 1 (COI) barcoding region from 144 gobioid fish species K2P, Kimura-2-parameter.
of the 144 species were the same. Every third codon base was highly variable with a 0.64 mean
R
seq
. The first (1.72) and the second codon bases (1.98) were nearly monomorphic. Nucleotide genetic distance parameters,
p
and K2P distance, also showed an almost zero rate of substitution for second nucleotide positions, with the first position being an order of magnitude higher and the third position being unparalleled among them (
Table 3
). The rate of synonymous substitutions was much higher than the rate of nonsynonymous substitutions (
Table 3
).
Three species could not be separated using the COI sequence analysis, including
Chaenogobius gulosus
(Gobionellinae),
Chaeturichthys stigmatias
(Gobionellinae), and
Lophiogobius ocellicauda
(Gobiinae). The
C. gulosus
sequence was highly similar to those of other congeneric species, as expected, whereas the
C. stigmatias
and
L. ocellicauda
results were very surprising. Multiple specimens of these species should be extensively analyzed in a future study to check the genetic divergence among these species; thus, those sequences were not deposited in GenBank. The average diversity among 142 haplotypes (
H
d
) was 0.999±0.001 (mean±standard deviation), and the average nucleotide diversity (π) was 0.199 ±0.002. Twenty-six genera were represented by two or more species. Levels of intragenus divergence were generally high (Table 4) but varied greatly among genera. For example, the average within-genus divergence of
Awaous, Eviota
, and
Trimma
was 29.85%, 29.13%, and 28.71%, respectively, which was larger than that of overall interspecific divergence. These values were considerably larger than those of
Amblygobius, Asterropteryx
and
Rhinogobius
(9.76%, 9.87%, and
Mean genetic divergences for the cytochrome oxidase c subunit 1 (COI) nucleotide sequences (p and Kimura-2-parameter [K2P] distances) among 144 gobioid speciesFour taxonomic levels are represented such as within-genus, within-subfamily, within-family, and among-family.
Mean genetic divergences for the cytochrome oxidase c subunit 1 (COI) nucleotide sequences (p and Kimura-2-parameter [K2P] distances) among 144 gobioid species Four taxonomic levels are represented such as within-genus, within-subfamily, within-family, and among-family.
10.74%, respectively). Mean interspecific divergences at higher taxonomic levels were slightly larger than that at the intragenus level, resulting in large overlaps among levels (
Table 4
). As roughly 90% of all gobiid fishes are either in Gobiinae or Gobionellinae, these two subfamilies were compared for nucleotide substitution rate (
Table 3
). Species from Gobiinae were consistently higher than Gobionelline fishes in every parameter estimated (
Table 3
).
Our phylogenetic data provide little evidence to support the previous claims at the generic and higher taxonomic levels, based on phenetic analyses. Several species did not cluster into their respective groups, and the BI tree failed to correctly identify some genera or subfamilies (
Fig. 1
). For example,
Tridentiger barbatus
(Gobiinellinae) clustered into the
Asterropteryx
clade (Gobiinae) with high nodal support rather than with other species of
Tridentiger. Pterogobius
zacalles
clustered with
Acanthogobius
, whereas
P. zonoleucus
formed a monophyletic group with the
Rhinogobius brun
-
Molecular phylogeny of Gobioidei based on Bayesian inference from 542 bp of the mitochondrial cytochrome oxidase c subunit 1 (COI) gene with two outgroup species from the Family Rhyacichthyidae: Protogobius attiti and Rhyacichthys guilberti. Under the chosen model, GTR+I+G, analysis was run for 10 million generations with sampling of one tree every 500 generations. Numbers above branches indicate posterior probabilities (¤0.8). Abbreviations in brackets on the right side indicate higher taxonomic names (subfamily and family) in current usage, including Amblyopinae (A), Benthophilinae (B), Gobiinae (I), Gobionellinae (N), Oxudercinae (O), Sicydiinae (S), Microdesmidae (MD), Odontobutidae (OD), and Rhyacichthyidae (RD). Asterisks immediately after the higher taxonomic names and arrows on nodes indicate taxa failing to resolve monophyletic assemblages, and species clustered into unrelated groups, respectively.
neus
complex, albeit without good nodal support. The BI tree also failed to resolve monophyletic assemblages of some taxa, such as
Awaous, Eviota, Trimma
, and Ondontobutidae. Those exceptions aside, the tree largely assigned species to identical major groups.
DISCUSSION
The COI sequences did resolve genealogical relationships well at the level of genera and family in Gobioidei. As previously noted in Che et al. (2012), this was possibly due to at least two factors. Most likely, the 542 nucleotides do not provide sufficient phylogenetically informative characters to recover the true phylogeny when examining hundreds of taxa with enormous diversity. In addition, the fast mutation rate and saturation in the third codon position can be a disadvantage at deeper phylogenetic levels, and the subsequent long terminal branches may impede resolution of ancient speciation due to the chance accumulation of shared character states (Huelsenbeck, 1997). Despite the poor monophyletic resolution in several taxa, some clear phylogenetic signatures were observed in the COI sequence data. For example, several major congeneric species,
Rhinogobius
and
Bathygobius
, tended to cluster together with no exception and, in most cases, so did consubfamilial species. It is believed that the utilization of more gene regions including nuclear DNA will assist in offering a more reliable phylogeny within Gobiidae and their placement within Gobioidei. Several nuclear genes such as recombination activating genes 1 and 2 (Rag 1 and 2) and ryandine receptor 3 (Ryr3) may have slower rate of sequence evolution in gobies compared to that of mtDNA genes (Yamada et al., 2009).
Our results support the validity of COI barcoding for species identification in gobiid species, although no attempt was made to include numerous specimens for any one species. One fundamental barcoding criterion is that congeneric divergence should be significantly higher than that of conspecific divergence (Hubert et al., 2008). The average intragenus distance (K2P) for 28 genera with multiple species in the present study was 21.09%, which was considerably higher than the values obtained among fish species in previous studies (9.93% from Ward et al., 2005; 9.54% from Ward and Holmes, 2007). In addition, the interspecific divergence at higher taxonomic levels was not significantly larger than that at the intragenus level, suggesting that congeneric gobies possess substantial interspecific sequence divergence in their COI genes. Significantly more nucleotide changes were observed at the third codon position than those at the first and the second, revealing that the extensive variation shown among the COI sequences typically reflects synonymous changes and little variation at the protein level. Consequently, the proportion of nonsynonymous to synonymous changes was far less than one (
Table 2
). As previously noted in Ward and Holmes (2007), this result must be due to exceptionally strong purifying selection of the COI gene and confirms that the ability of COI to identify species in Gobiidae is dependent on the degenerate nature of the genetic code.
We did not provide sufficient evidence for the utility of the COI towards cryptic species identification in several species complexes. Gobiidae taxonomy has been studied extensively for the last several decades, but confusion still exists. One typical case is the taxonomic status of the Gymnogobius species complex (
G. urotaenia, G. opperiens
, and
G. petschiliensis
) (e.g., Harada et al., 2002). In our results, overall K2P divergence within
Gymnogobius
was 12.74%, whereas the average value among the
Gymnogobius
species complex was only 1.81%, probably reflecting a short history of reproductive isolation. The
Rhinogobius brunneus
complex (in our analysis,
Rhinogobius brunneus, R. brunneus
CB, and
R. brunneus
CO) is also a representative example of a cryptic species complex in Gobiidae (Kawanabe and Mizuno, 1989; Kim, 1995). Overall, K2P divergence within
Rhinogobius was
>10%, whereas the average value among the
R. brunneus
complex was just 2%. Although more work needs to be done with multiple specimens, the COI sequence may not be a reliable tool to delineate cryptic and complex species boundaries in the family Gobiidae.
Acknowledgements
We thank Mr. Seong Jang Jo and Jin Young Choo for theirassistance collecting fish. We also thank Seongjun Park forlogistical phylogenetic analyses support. Su Youn Baekprovided assistance with the laboratory studies. This studywas supported by the grant “Korean Tree of Life, 4th Year:Gobioidei Fishes (No. 211C000068)” funded by the NationalInstitute of Biological Resources, Korean Government andawarded to H. Y. Suk.
Akihito A
,
Iwata T
,
Kobayashi T
,
Ikeo K
,
Imanishi T
,
Ono H
,
Umehara Y
,
Hamamatsu C
,
Sugiyama K
,
Ikeda Y
,
Sakamoto K
,
Fumihito A
,
Ohno S
,
Gojobori T
2000
Evolutionaryaspects of gobioid fishes based upon a phylogenetic analysisof mitochondrial cytochrome b genes.
Gene
259
5 -
15
April J
,
Mayden RL
,
Hanner RH
,
Bernatchez L
2011
Geneticcalibration of species diversity among North America’sfreshwater fishes.
Proceedings of the National Academy of Sciences United States of America
108
10602 -
10607
Aquilino SVL
,
Tango JM
,
Fontanilla IKC
,
Pagulayan RC
,
Basiao ZU
,
Ong PS
,
Quilang JP
2011
DNA barcoding of the ichthyofaunaof Taal Lake, Philippines.
Molecular Ecology Resources
11
612 -
619
Aquino LMG
,
Tango JM
,
Canoy RJC
,
Fontanilla IKC
,
Basiao ZU
,
Ong PS
,
Quilang JP
2011
DNA barcoding of fishes ofLaguna de Bay, Philippines.
Mitochondrial DNA
22
143 -
153
Avise JC
,
Ellis D
1986
Mitochondrial DNA and the evolutionarygenetics of higher animals.
Philosophical Transactions of the Royal Society of London Series B Biological Sciences
312
325 -
342
Brown WM
,
George M Jr
,
Wilson AC
1979
Rapid evolutionof animal mitochondrial DNA.
Proceedings of the National Academy of Sciences of the United States America
76
1967 -
1971
Che J
,
Chen HM
,
Yang JX
,
Jin JQ
,
Jiang K
,
Yuan ZY
,
Murphy RW
,
Zhang YP
2012
Universal COI primers for DNAbarcoding amphibians.
Molecular Ecology Resources
12
247 -
258
Clare EL
,
Lim BK
,
Engstrom MD
,
Eger JL
,
Hebert PDN
2007
DNA barcoding of Neotropical bats: species identificationand discovery within Guyana.
Molecular Ecology Notes
7
184 -
190
Ebach MC
,
Holdrege C
2005
More taxonomy, not DNA barcoding.
BioScience
55
822 -
823
Feng Y
,
Li Q
,
Kong L
,
Zheng X
2011
COI-based DNA barcodingof Arcoida species (Bivalvia: Pteriomorphia) along thecoast of China.
Molecular Ecology Resources
11
435 -
441
Hall TA
1999
BioEdit: a user-friendly biological sequencealignment editor and analysis program for Windows 95/98/NT.
Nucleic Acids Symposium Series
41
95 -
98
Harada S
,
Jeon SR
,
Kinoshita I
,
Tanaka M
,
Nishda M
2002
Phylogenetic relationships of four species of floating gobies(Gymnogobius) as inferred from partial mitochondrial cytochromeb gene sequences.
Ichthyological Research
49
324 -
332
Hebert PDN
,
Cywinska A
,
Ball SL
,
DeWaard JR
2003
Biologicalidentifications through DNA barcodes.
Proceedings of the Royal Society B: Biological Sciences
270
313 -
321
Hebert PDN
,
Stoeckle MY
,
Zemlak TS
,
Francis CM
2004
Identification of birds through DNA barcodes.
PLoS Biology
2
e312. -
Herler J
,
Munday PL
,
Hernaman V
2011
Gobies on coral reefs.In: The biology of gobies (Eds., Patzner RA, van TassellJL, Kovačić M, Kapoor BG).
Science Publishers
New York
493 -
530
Hubert N
,
Hanner R
,
Holm E
,
Mandrak NE
,
Taylor E
,
Burridge M
,
Watkinson D
,
Dumont P
,
Curry A
,
Bentzen P
,
Zhang J
,
April J
,
Bernatchez L
2008
Identifying Canadian freshwaterfishes through DNA barcodes.
PLoS ONE
3
e2490 -
Hubert N
,
Meyer CP
,
Bruggemann HJ
,
Guérin F
,
Komeno RJL
,
Espiau B
,
Causse R
,
Williams JT
,
Planes S
2012
Crypticdiversity in Indo-Pacific coral-reef fishes revealed by DNAbarcodingprovides new support to the centre-of-overlap hypothesis.
PLoS ONE
7
e28987 -
Huelsenbeck JP
1997
Is the Felsenstein zone a fly trap?
Systematic Biology
46
69 -
74
Ivanova NV
,
Zemlak TS
,
Hanner RH
,
Hebert PDN
2007
Universalprimer cocktails for fish DNA barcoding.
Molecular Ecology Notes
7
544 -
548
Johns GC
,
Avise JC
1998
A comparative summary of geneticdistances in the vertebrates from the mitochondrial cytochromeb gene.
Molecular Biology and Evolution
15
1481 -
1490
Kawanabe H
,
Mizuno N
1989
Freshwater fishes of Japan.
Yamatokeikokusha
Tokyo
1 -
719
Keith P
,
Lord C
,
Lorion J
,
Watanabe S
,
Tsukamoto K
,
Couloux A
,
Dettai A
2011
Phylogeny and biogeography of Sicydiinae(Teleostei: Gobiidae) inferred from mitochondrial andnuclear genes.
Marine Biology
158
311 -
326
Kim JB
1995
The studies of speciation and systematics on thefishes of the genera Rhinogobius and Tridentiger (Perciformes,Gobiidae) in Korea.
Inha University
Incheon, Korea
PhD dissertation
1 -
158
Kimura M
1980
A simple method for estimating evolutionaryrates of base substitutions through comparative studies ofnucleotide sequences.
Journal of Molecular Evolution
16
111 -
120
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
Larson HK
2001
A revision of the gobiid fish genus Mugilogobius(Teleostei: Gobioidei), and its systematic placement.
Records of the Western Australian Museum, Supplement
62
1 -
233
Leray M
,
Boehm JT
,
Mills SC
,
Meyer CP
2012
Moorea BIOCODEbarcode library as a tool for understanding predatorpreyinteractions: insights into the diet of common predatorycoral reef fishes.
Coral Reefs
31
383 -
388
Moore WS
1995
Inferring phylogenies from mtDNA variation:mitochondrial-gene trees versus nuclear-gene trees.
Evolution
49
718 -
726
Murdy EO
1989
A taxonomic revision and cladistic analysisof the oxudercine gobies (Gobiidae: Oxudercinae).
Records of the Australian Museum, Supplement
11
1 -
93
Nei M
,
Gojobori T
1986
Simple methods for estimating thenumbers of synonymous and nonsynonymous nucleotidesubstitutions.
Molecular Biology and Evolution
3
418 -
426
Nei M
,
Kumar S
2000
Molecular evolution and phylogenetics.
Oxford University Press
Oxford
1 -
352
Nelson JS
2006
Fishes of the world.
4th ed
John Wiley and Sons
Hoboken, NJ
1 -
601
Parenti LR
,
Thomas KR
1998
Pharyngeal jaw morphologyand homology in sicydiine gobies (Teleostei: Gobiidae) andallies.
Journal of Morphology
237
257 -
274
Pezold F
1993
Evidence for a monophyletic Gobiinae.
Copeia
1993
634 -
643
Posada D
,
Crandall KA
1998
Modeltest: testing the model of DNA substitution.
Bioinformatics
14
817 -
818
Ronquist F
,
Huelsenbeck JP
2003
MRBAYES 3: Bayesianphylogenetic inference under mixed models.
Bioinformatics
19
1572 -
1574
Roques S
,
Fox CJ
,
Villasana MI
,
Rico C
2006
The completemitochondrial genome of the whiting, Merlangius merlangusand the haddock, Melanogrammus aeglefinus: a detailedgenomic comparison among closely related species of theGadidae family.
Gene
383
12 -
23
Steinke D
,
Zemlak TS
,
Hebert PDN
2009
Barcoding nemo:DNA-based identifications for the ornamental fish trade.
PLoS ONE
4
e6300 -
Tamura K
,
Peterson D
,
Peterson N
,
Stecher G
,
Nei M
,
Kumar S
2011
MEGA5: Molecular Evolutionary Genetics Analysisusing maximum likelihood, evolutionary distance, andmaximum parsimony methods.
Molecular Biology and Evolution
28
2731 -
2739
Thacker CE
2003
Molecular phylogeny of the gobioid fishes(Teleostei: Perciformes: Gobioidei).
Molecular Phylogenetics and Evolution
26
354 -
368
Thacker CE
2009
Phylogeny of Gobioidei and placement withinAcanthomorpha, with a new classification and investigationof diversification and character evolution.
Copeia
93
Thacker CE
,
Hardman MA
2005
Molecular phylogeny of basalgobioid fishes: Rhyacichthyidae, Odontobutidae, Xenisthmidae,Eleotridae (Teleostei: Perciformes: Gobioidei).
Molecular Phylogenetics and Evolution
37
858 -
871
Thacker CE
,
Schaefer SA
2000
Phylogeny of the wormfishes(Teleostei: Gobioidei: Microdesmidae).
Copeia
2000
940 -
957
Thacker CE
,
Thompson AR
,
Roje DM
2011
Phylogeny andevolution of Indo-Pacific shrimp-associated gobies (Gobiiformes:Gobiidae).
Molecular Phylogenetics and Evolution
59
168 -
176
Tornabene L
,
Baldwin C
,
Weigt LA
,
Pezold F
2010
Exploringthe diversity of western Atlantic Bathygobius (Teleostei:Gobiidae) with cytochrome c oxidase-I, with descriptions of two new species.
Aqua Journal of Ichthyology and Aquatic Biology
16
141 -
170
Triantafyllidis A
,
Bobori D
,
Koliamitra C
,
Gbandi E
,
Mpanti M
,
Petriki O
,
Karaiskou N
2011
DNA barcoding analysisof fish species diversity in four north Greek lakes.
Mitochodrial DNA
22
37 -
42
Wang HY
,
Tsai MP
,
Dean J
,
Lee SC
2001
Molecular phylogenyof gobioid fishes (Perciformes: Gobioidei) based onmitochondrial 12S rRNA sequences.
Molecular Phylogenetics and Evolution
20
390 -
408
Ward RD
,
Holmes BH
2007
An analysis of nucleotide andamino acid variability in the barcode region of cytochrome coxidase I (cox1) in fishes.
Molecular Ecology Notes
7
899 -
907
Ward RD
,
Zemlak TS
,
Innes BH
,
Last PR
,
Hebert PDN
2005
DNA barcoding Australia’s fish species.
Philosophical Transactions of the Royal Society of London Series B Biological Sciences
360
1847 -
1857
Weigt LA
,
Baldwin CC
,
Driskell A
,
Smith DG
,
Ormos A
,
Reyier EA
2012
Using DNA barcoding to assess Caribbean reeffish biodiversity: expanding taxonomic and geographiccoverage.
PLoS ONE
7
e41059 -
Will KW
,
Mishler BD
,
Wheeler QD
2005
The perils of DNAbarcoding and the need for integrative taxonomy.
Systematic Biology
54
844 -
851
Yamada T
,
Sugiyama T
,
Tamaki N
,
Kawakita A
,
Kato M
2009
Adaptive radiation of gobies in the interstitial habitats ofgravel beaches accompanied by body elongation and excessivevertebral segmentation.
BMC Evolutionary Biology
9
145 -
Zander CD
2011
Morphological adaptation to special environmentsof gobies. In: The biology of gobies (Eds., PatznerRA, van Tassell JL, Kovačić M, Kapoor BG).
Science Publishers
New York
345 -
366