Clarifying the taxonomic status of Merluccius spp. in the northeastern Pacific: A combined morphological and molecular approach

Abstract

The taxonomic status of hake (Merluccius spp.) in the northeastern Pacific is unclear. Hakes in this region are Merluccius productus, M. angustimanus, M. hernandezi, and a morphotype known as dwarf hake. Of these, only the first two species are currently valid. Descriptions in previous
studies have been limited by overlapping morphological characteristic, lack of biological material, and limited numbers of
sampling localities. To clarify their taxonomy, 461 hake were obtained from eight localities along the North American coast
for morphological and mitochondrial DNA sequence analyses (cytochrome b, cytochrome c oxidase subunit I, and 16S ribosomal rDNA). Morphological and molecular analyses suggest that hake in this region represent
a continuum of a single species with some levels of morphological and genetic intra-specific variation. In light of these
results, we propose that M. productus is the only species of hake present along the North American and northern Central American coast.

Keywords
Merluccius
–Northeastern Pacific–Morphology–Cytochrome b
–Cytochrome c oxidase subunit I–16S rDNA

Full text

RESEARCH PAPER
Clarifying the taxonomic status of Merluccius spp.
in the northeastern Pacific: a combined morphological
and molecular approach
Claudia A. Silva-Segundo •Mariela Brito-Chavarria •Eduardo F. Balart •
Irene de los A. Barriga-Sosa •Roberto Rojas-Esquivel •
Marı
´a Ine
´s Rolda
´n•Gopal Murugan •Francisco J. Garcı
´a-De Leo
´n
Received: 26 November 2009 / Accepted: 4 May 2010 / Published online: 28 May 2010
ÓSpringer Science+Business Media B.V. 2010
Abstract The taxonomic status of hake (Merluccius
spp.) in the northeastern Pacific is unclear. Hakes in
this region are Merluccius productus,M. angustim-
anus,M. hernandezi, and a morphotype known as
dwarf hake. Of these, only the first two species are
currently valid. Descriptions in previous studies have
been limited by overlapping morphological charac-
teristic, lack of biological material, and limited
numbers of sampling localities. To clarify their
taxonomy, 461 hake were obtained from eight
localities along the North American coast for mor-
phological and mitochondrial DNA sequence analy-
ses (cytochrome b, cytochrome coxidase subunit I,
and 16S ribosomal rDNA). Morphological and
molecular analyses suggest that hake in this region
represent a continuum of a single species with some
levels of morphological and genetic intra-specific
variation. In light of these results, we propose that
M. productus is the only species of hake present
along the North American and northern Central
American coast.
Keywords Merluccius Northeastern Pacific 
Morphology Cytochrome bCytochrome c
oxidase subunit I 16S rDNA
Introduction
The genus Merluccius (hake) is currently represented
by 12 species that are widely distributed in the
Atlantic, Pacific, and around New Zealand (Inada
1981). Hake inhabit deep waters over the continental
shelf and exhibit diurnal vertical movement and
seasonal horizontal migration influenced by feeding
and reproductive behavior (Inada 1981; Lloris et al.
2003). The taxonomy of Merluccius is complex; in
fact, identification of species is difficult because of
their similarities in external morphology (Ginsburg
1954; Lozano-Cabo 1965; Inada 1981; Cohen et al.
1990; Lloris et al. 2003). Coexistence of hake species
C. A. Silva-Segundo G. Murugan F.
J. Garcı
´a-De Leo
´n(&)
Laboratorio de Gene
´tica para la Conservacio
´n, Centro de
Investigaciones Biolo
´gicas del Noroeste, Mar Bermejo
195, Col. Playa Palo de Santa Rita, 23090 La Paz, BCS,
Me
´xico
e-mail: fgarciadl@cibnor.mx
M. Brito-Chavarria E. F. Balart
Laboratorio de Necton & Coleccio
´n Ictiolo
´gica, Centro de
Investigaciones Biolo
´gicas del Noroeste, Mar Bermejo
195, Col. Playa Palo de Santa Rita, 23090 La Paz, BCS,
Me
´xico
I. de los A. Barriga-Sosa R. Rojas-Esquivel
Laboratorio de Gene
´tica y Biologı
´a Molecular, Planta
Experimental de Produccio
´n Acuı
´cola, Departamento de
Hidrobiologı
´a, Divisio
´n de Ciencias Biolo
´gicas y de la
Salud, 9340 Me
´xico City, Mexico
M. I. Rolda
´n
Laboratorio de Ictiologı
´a Gene
`tica, Departamento de
Biologı
´a, Universidad de Girona, Campus Montilivi,
17071 Girona, Spain
123
Rev Fish Biol Fisheries (2011) 21:259–282
DOI 10.1007/s11160-010-9166-6

is common and they are typically larger in size at
higher latitudes than in the tropics (Inada 1981;
Cohen et al. 1990). These conditions conceivably
contribute to the difficulty of species identification.
In the northeastern Pacific (west coast of North
America), three species of Merluccius have been
described: M. productus (Ayres 1855), distributed
from the northern part of Vancouver Island, Canada to
the Gulf of Tehuantepec, Mexico; M. angustimanus
(Garman 1899), found from the Gulf of California,
Mexico to Ensenada de Tumaco (1.7°N, 78.3°W),
Colombia (Lloris et al. 2003); and M. hernandezi
(Mathews 1985), present only in the Upper Gulf of
California, Mexico. Previous studies have also
described a morphotype called ‘‘dwarf hake’’ off the
west coast of the State of Baja California Sur, which
has been assigned to M. productus (Bailey et al. 1982).
Some authors have argued that ‘‘dwarf hake’’ is
M. angustimanus, based on similarities in morphologies
and allozyme genotypes (Vrooman and Paloma 1977;
Cohen et al. 1990;Inada1995;Balart-Pa
´ez 2005). In a
review of hakes of the world (Family Merlucciidae)
Lloris et al. (2003)proposedM. hernandezi asynonym
of M. angustimanus because there is high variability
and overlap in meristic characteristics.
Taxonomic uncertainty of the northeastern Pacific
hake is further complicated by the lack of information
on morphological variability in M. angustimanus
throughout its range, since the original description
was based on a small number of specimens from the
Gulf of Panama (Garman 1899) and no comparisons
were made with other species of hake. Later studies
incorporated samples collected off California and
southern Baja California (Ginsburg 1954; Mathews
1985; Inada 1995; Balart-Pa
´ez 2005); the morpholog-
ical variability of M. angustimanus between this area
and Panama is unknown. Likewise, studies of ichthy-
oplankton have failed to detect the presence of
M. angustimanus larvae off California and Baja
California and only M. productus has been found
(Ahlstrom and Counts 1955; Ahlstrom 1969; Ambrose
1996; Smith and Moser 2003). This is puzzling
because adult M. angustimanus have been reported
in this region (Lloris et al. 2003). Possible causes could
be related to foraging migration of adult M. angustim-
anus and reproduction or incorrect identification.
Additionally, when comparing all the taxonomic
descriptions used for identification of hake along the
west coast, there is a broad overlap of morphological
characteristics. This overlap is evident in the number
of fin rays, number of gill rakers, number of
vertebrae, and head length expressed as percent of
standard body length (Ayres 1855; Garman 1899;
Ginsburg 1954; Vrooman and Paloma 1977; Inada
1981; Mathews 1985; Balart-Pa
´ez 2005; Table 1).
The limited knowledge of morphological and
genetic variability of hake along this coast has also
hindered comprehensive understanding of their phy-
logenetic relationships and geographic distribution.
Biogeographic and phylogenetic studies have estab-
lished that hake are a monophyletic group that
originated in the North Atlantic during the Oligocene;
during the Miocene, it split into two main clades
(Inada 1981; Kabata and Ho 1981;Ho1990; Rolda
´n
et al. 1999; Quinteiro et al. 2000; Grant and Leslie
2001; Campo et al. 2007). The Euro-African clade
consisting of hake from the coasts of Western Europe
and Western Africa (M. merluccius,M. capensis,
M. senegalensis,M paradoxus, and M. polli) and an
American clade that includes hake from the Atlantic
(M. bilinearis,M. albidus and M. hubbsi) and Pacific
(M. gayi,M. productus,M. angustimanus and M. australis)
sides of North America. The relationships among
species of the American clade is complex and has not
been convincingly resolved because there is insuffi-
cient reliable genetic information for some species,
including the hake that inhabit the west coast, such as
M. angustimanus and M. hernandezi (Grant and
Leslie 2001). Additionally, geological events, includ-
ing the closure of the Panama Seaway by the
formation of the Isthmus of Panama over 3 million
years ago (Bermingham and Lessios 1993; Knowlton
et al. 1993; Bermingham et al. 1997) and the reopening
of the marine peninsular corridor 1–1.6 million years
ago in the Baja California Peninsula (Beal 1948;
Riddle et al. 2000; Bernardi et al. 2003), have shaped
the phylogeography of west coast hake (Inada 1981;
Ho 1990).
Historically, morphological analyses have been
used for taxonomic studies. Combining these types of
analyses with genetic studies provides a more robust
foundation to make taxonomic decisions. Sequences
of mitochondrial DNA are useful as molecular
markers for identifying species and determining the
genetic relationships of many fish taxa (Ovenden
1990; Billington and Hebert 1991; Meyer 1993). In
particular, cytochrome b, cytochrome coxidase
subunit I, and 16S ribosomal rDNA have been widely
260 Rev Fish Biol Fisheries (2011) 21:259–282
123

Table 1 Meristic and morphometric data of taxonomic descriptions used for species differentiation of northeastern Pacific hake (Merluccius productus,M. angustimanus,
M. hernandezi and the morphotypes ‘‘dwarf hake’’)
Species Author Number of rays Number of gill rakers NV HL/SL Collection area n
1D 2D A RPec Upper Lower Total
M. productus Ayres (1855) 11 40 40 16 – – – – – San Francisco, CA –
Ginsburg (1954) 10–12 39–44 41–44 14–16 4–5 13–18 17–22 – 25.5–28.5
(27.27)
WA, OR, and CA 12
Vrooman and
Paloma (1977)
– 37–42
(40.9)
37–44
(41)
– 3–5
(4.3)
13–18
(15.1)
16–22
(19.4)
50–55
(52.7)
25.6–33.3
(28.7)
CA and Baja California 105
Inada (1981) 9–12
(10.3)
39–44
(40.9)
39–44
(41.1)
14–17
(15.6)
3–6
(4.4)
14–17
(15.6)
17–23
(20.1)
53–54
(53.4)
24.7–28.9
(26.3)
WA, OR, and CA 70
M. angustimanus Garman (1899) 10 35 36 17 4 14 18 – – Gulf of Panama
´–
Ginsburg (1954) 11–12 36–39 37–40 15–17 3–5 12–14 16–18 – 31–33.5
(32.3)
CA, Gulf of California,
and Gulf of Panama
8
Inada (1981) 9–12
(9.8)
36–40
(37.6)
36–39
(37.3)
14–17
(16)
3–5 (4) 12–14
(13.1)
16–18
(17.1)
49–52
(50.4)
30.1–33.5
(31.9)
Baja California 45
Mathews (1985) – 34–39
(36.7)
32–40
(36)
– – – 14–18
(16.1)
45–49
(47)
26.1–29.7
(27.3)
Sinaloa and Sonora 83
Balart-Pa
´ez (2005) 9–12
(10.2)
35–41
(38.2)
34–42
(38.7)
– – 10–15
(13.4)
– – 27.2–47.6
(32.4)
Baja California 81
M. hernandezi Mathews (1985) – 36–42
(39.5)
37–42
(39.8)
– – 14–20
(17.3)
47–52
(49.1)
23.7–30.5
(25.6)
Upper Gulf of California 159
Dwarf hake Vrooman and
Paloma (1977)
– 35–41
(37.5)
35–41
(37.7)
– 3–5
(3.9)
11–16
(13.4)
14–20
(17.3)
47–53
(50.5)
26.5–35.4
(31.25)
Baja California 231
1D =first dorsal fin rays, 2D =second dorsal fin rays, A =anal fin rays, RPec =pectoral fin rays, HL/SL =head length expressed as percent of standard body length,
NV =number of vertebrae, n=sample size
Rev Fish Biol Fisheries (2011) 21:259–282 261
123

used for analysis and comparison of taxonomic
groups, where the degree of genetic distance between
the nucleotide sequences can be used to determine
divergence within and among species (Johns and
Avise 1998; Nei and Kumar 2000; Astrin et al. 2006;
Kartavtsev and Lee 2006; Kartavtsev et al. 2007).
Development of a commercial hake fishery in
Mexican waters will require solving international
regulatory initiatives among Canada, the United
States, and Mexico; this includes assessment of
various stocks in the region; thus, it is critical to have
tools to efficiently identify species. This study
addresses the taxonomic uncertainty of hake species
along the western coasts of North and Central America
by collecting a large sample of specimens represent-
ing, for the first time, a very larger geographical
distribution of hake species and simultaneously apply-
ing two fundamental tools for identifying and differ-
entiating among the species: (1) A study using
morphometric and meristic characteristics; and (2) A
study using three mitochondrial DNA genes: cyto-
chrome b(Cyt-b), cytochrome c oxidase subunit I
(COI), and 16S ribosomal rDNA (16S).
Materials and methods
Sampling
Samples of hake from Mexican waters were obtained
from two fishing cruises: one in the Upper Gulf of
California (UG) on the research vessel ‘BIP XI’ of
the Centro Regional de Investigacio
´n Pesquera-Insti-
tuto Nacional de la Pesca (CRIP-INP); the second in
the southwest coast of the peninsula of Baja Califor-
nia near Bahia Sebastian Vizcaino (VIZ) and the
southern part of Baja California Sur state (BCS)
aboard the Research Vessel ‘BIP XII’, of the Centro
de Investigaciones Biolo
´gicas del Noroeste (CIB-
NOR). Samples from U.S. waters came from the
research fishing cruise carried out in Washington
(W), Oregon (O) and San Francisco, California (SF)
by research vessels of the National Oceanic and
Atmospheric Administration. Additional samples
were taken by commercial fishing fleets from Eureka,
California (E) and Costa Rica (CR) (Fig. 1; Table 2).
A total of 461 adult hakes were collected from North
Pacific coast and frozen prior to their morphological
analysis, then preserved in isopropyl alcohol (50%).
Muscle and gills samples for DNA extraction of 110
specimens were preserved in 96% alcohol. Both, all
fishes and tissue samples were deposited in the fish
collection (catalog numbers: CIBN4360–CIBN4383)
and collection of the Laboratory of Conservation
Genetics at CIBNOR, respectively.
Morphological analysis
A first approach to identify the collected organisms
was conducted by comparing the diagnostic morpho-
logical characteristics used in previous taxonomic
studies to separate putative species M. angustimanus
and M. productus. These characteristics were: num-
ber of rays of the first dorsal fin rays, second dorsal
fin, pectoral fin, and anal fin; number of gill rakers of
the first branchial arch; and head length relative to the
standard length as described by many authors (Ayres
1855; Garman 1899; Ginsburg 1954; Vrooman and
Paloma 1977; Inada 1981,1995; Mathews 1985;
Cohen et al. 1990; Lloris et al. 2003; Balart-Pa
´ez
2005; Table 1).
Additionally, a principal component analysis
(PCA) was carried out to assess patterns of morpho-
logical variation between all specimens (461
120°110°100°90°80°
Longitude (°W)
10°
20°
30°
40°
50°
Latitude (°N)
W
O
SF
UG
VIZ
BCS
UNITED STATES OF AMERICA
MEXICO
CR
N
0°2°4°6°8°10°
0600 Nm
E
Fig. 1 Collection areas of hake (Merluccius spp.) along the
west coast of North and Central America: Washington (W),
Oregon (O), Eureka, California (E), San Francisco, California
(SF), Upper Gulf of California (UG), Bahı
´a Sebastian
Vizcaino, Baja California (VIZ), the southern part of the State
of Baja California Sur (BCS), and Costa Rica (CR)
262 Rev Fish Biol Fisheries (2011) 21:259–282
123

individuals) collected and identify the morphological
characteristics that determine the formation of groups
within the multivariate space (Crisci and Lo
´pez-
Armengol 1983; Quinn and Keough 2002). For this
analysis, we recorded the meristic and morphometric
characteristics suggested in the literature (Ginsburg
1954; Hubbs and Lagler 1958). The meristic charac-
teristics used were: the number of first dorsal fin rays
(1D), number of second dorsal fin rays (2D), number
of pectoral fin rays (RPec), number of anal fin rays
(A), and number of gill rakers of the first branchial
arch (GRI). Morphometric measures were expressed
as percentages of head length (HL), standard length
(SL), and diameter of eye orbit (DO). HL-based
characteristics were: length of pectoral fin (LPec/
HL), length of the upper jaw (ML/HL), pre-orbital
length (POL/HL), diameter of orbit (DO/HL), inter-
orbital distance (IO/HL); SL-based characteristics
were: head length (HL/SL), length of pectoral fin
(LPec/SL), length of the upper jaw (ML/SL), pre-
orbital length (POL/SL), diameter of orbit (DO/SL),
length from tip of snout to pectoral fin insertion
(S-Pec/SL), length from tip of the snout to anal fin
origin (S-A/SL). The one-DO based characteristic we
measured was inter-orbital distance (IO/DO). All
characteristics for PCA were standardized following
the method of Legendre and Legendre (1998).
A third analysis was carried out only with the
characteristics that explained the highest percentage
of variance in the PCA (eigenvalues C0.6). These
characteristics were used for graphic comparisons
following a method similar to Hubbs and Hubbs
(1953), which uses the mean, range, and 95%
confidence intervals of individual characteristics. To
verify the existence of clinal or allometric effects on
these characteristics, multiple linear regressions were
performed. These regressions used collection site
latitude and specimen standard length as independent
variables. For all morphological analyses, we used
Statistica 7.0 (StatSoft 2004).
Even though otolith morphometrics have proved
useful for identifying species (Frost 1981; Short et al.
2006), in this study, only limited otolith data were
included for the regression analysis because many
were damaged. The otoliths (sagittae) were extracted
by cutting at the level of the otic capsule, after
removing the branchial apparatus, and digitized with
600 dpi resolution with a scanner. Maximum width
(WO) and length (LO) of each right otolith was
measured using the SigmaScan Pro 5 software. If the
right otolith was broken, the left sagittae was
measured.
DNA extraction, amplification, and sequencing
DNA was extracted with a Dneasy
TM
Tissue Kit
(Qiagen, Hilden, Germany) and sodium chloride as
described by Aljanabi and Martinez (1997). PCR
amplification for Cyt-bwas performed with primers
L14322 and H14747 described by Teletchea et al.
(2006); for COI, with L6154 and H6556 (Teletchea
et al. 2006); and for 16S, 16S ar-L and 16S br-H
(Kocher et al. 1989; Huelsenbeck et al. 1996). PCR
reactions were carried out in a solution containing
50 ll of 90 ng total DNA, 19buffer, 3 mMMgCl
2
,
0.2 mM of each dNTP, 0.5 lM of each primer, and
0.75 U Taq polymerase. PCR was performed in a
thermocycler (Bio-Rad DNA Engine Peltier) under
the following conditions: initial denaturing step at
94°C for 5 min, 30 cycles of denaturing at 94°C for
Table 2 Collection sites, codes, date, origin, geographic coordinates, and sample size (n) used in morphological and genetic analysis
(parenthesis)
Location Code Date Origin Latitude N Longitude W n
Washington, USA W Jul 2007 NOAA 47°55.42 125°49.20 16 (17)
Oregon, USA O Jul 2007 NOAA 42°56.20 124°46.40 15 (16)
Eureka, CA, USA E May 2006 Commercial 40°46.30 124°28.13 25 (4)
San Francisco, CA, USA SF Jul 2007 NOAA 37°36.90 122°57.12 15 (18)
Upper Gulf of California, MX UG Nov 2006, Apr 2007 CRIP-INP 29807.98 112837.07 238 (23)
Bahı
´a Sebastia
´n Vizcaı
´no,
Baja California, MX
VIZ Mar 2006 CIBNOR 28°37.49 114°53.49 30 (10)
Baja California Sur, MX BCS Mar 2006 CIBNOR 23°29.82 110°36.41 114 (14)
Costa Rica CR Nov 2007 Commercial 9°33.52 84°49.24 8 (8)
Rev Fish Biol Fisheries (2011) 21:259–282 263
123

45 s, annealing (for 45 s) at 50°C for Cyt-band COI
and 53°C for 16S, followed by an extension at 72°C
for 1.5 min, ending with a final extension at 72°C for
10–15 min. For purification of PCR products, we
used ExoSap-It and QIAquick Kits (Qiagen). DNA
sequencing reactions were carried out following the
manufacturer’s instructions (Big Dye Terminator
cycle sequencing kit, PerkinElmer, Waltham, MA).
Primers used for sequencing were the same as for
PCR. Capillary electrophoresis was performed with a
genetic analyzer (ABI PRISM 3100 Avant Genetic
Analyzer) at the Laboratorio Divisional de Biologı
´a
Molecular, Universidad Auto
´noma Metropolitana,
Unidad Iztapalapa, Mexico and with a genetic
analyzer (ABI PRISM 3730XL, Macrogen, Seoul,
Korea).
Sequence data analysis
The nucleotide sequences of the three amplified
genes (Cyt-b, COI, and 16S) were edited and aligned
using MEGA 4.0 (Tamura et al. 2007). Alignment
was performed with Clustal W (Thompson et al.
1994) with default values (gap penalty 15 and gap
extension 6.66) included in MEGA. Observed align-
ment was clean and without gaps in Cyt-b, COI, and
16S datasets. Cyt-band COI sequences were trans-
lated into amino acid sequences as an additional
check of alignment using MEGA. Nucleotide satura-
tion was tested with DAMBE version 5.0.45 (Xia and
Xie 2001). A partition homogeneity test was imple-
mented in PAUP version 4.0b10 (Swofford 2003)to
determine whether to combine or not combine the
three gene datasets. They were compatible
(P=0.96), as one would expect for three mitochon-
drial partitions. Thus, we defined a combined frag-
ment consisting of the three genes (Cyt-b, COI, and
16S). Subsequently for the individual genes and the
combined fragment, we estimated nucleotide diver-
sity (p) and haplotype (h) diversity using DnaSP 4.0
(Rozas et al. 2003). The haplotypes were confirmed
by sequencing in the opposite direction to corroborate
nucleotide composition. All haplotypes were depos-
ited in GenBank: Accession numbers for Cyt-b
haplotypes are GQ503215–GQ503232; COI haplo-
types are GQ503196–GQ503214; and 16S haplotypes
are GQ503180–GQ503195.
Genetic divergence between the haplotypes found
in this study was estimated using the Tamura and Nei
(1993) model implemented in MEGA version 4.0
(Tamura et al. 2007). To determine the level of
genetic divergence between our haplotypes and those
of other hake species, we retrieved homologous
sequences registered in GenBank (Table 3). These
sequences were aligned (without gaps) with our
datasets for the three individual genes and the
combined fragment to compare genetic distances
within and between hake species. To compare COI
and the combined fragment, only sequences for four
species of hake were available from GenBank
(M. merluccius,M. bilinearis,M. gayi, and M. pro-
ductus; Table 3). To determine whether inter- and
intra-specific differences could be detected within the
genus Merluccius and define the level of haplotype
divergence in the samples, three comparison groups
were defined: (1) Haplotypes, distances between the
haplotypes determined in this study: (2) Within-
species distances (GenBank sequences); and (3)
Between-species distances (GenBank sequences).
Subsequently, the genetic distances (Tamura-Nei)
were plotted and an ANOVA test was carried out to
measure significant distance differences between the
groups with Statistica 7.0 (StatSoft 2004).
Phylogenetic inference was conducted with the
combined fragment (Cyt-b?COI ?16S) and using
neighbor-joining (Saitou and Nei 1987) and Bayesian
analyses (Huelsenbeck and Ronquist 2001). The
neighbor-joining analysis was conducted in MEGA
using the Tamura-Nei model (Tamura and Nei 1993)
as the best-fit model by ModelTest 3.5 (Posada and
Crandall 1998). The statistical robustness was tested
with 5000 bootstrap replicates (Felsenstein 1985).
The Bayesian analysis was performed using MrBayes
3.1.1 with default settings to establish the initial
heating values for four Markov chains, which ran
simultaneously and sampled [1 million generations
with sampling every 100 generations. Stability was
determined by plotting likelihoods against generation
model; 5,000 trees were excluded as burn-in. The
remaining post-burn-in trees were used to construct a
50% majority-rule consensus tree and posterior
probabilities assigned for each node. MrModelTest
version 2.2 (Nylander 2004) was employed to
determine the model of sequence evolution that best
fitted our dataset and be included in MrBayes. A
combined sequence of M. bilinearis, the most diver-
gent taxon among the American species (Rolda
´n et al.
1999; Quinteiro et al. 2000; Grant and Leslie 2001;
264 Rev Fish Biol Fisheries (2011) 21:259–282
123

Campo et al. 2007) retrieved from GenBank, was
included as outgroup; and combined sequences of
M. gayi and M. productus, American species from the
west coast of the southern and northern East Pacific,
respectively, retrieved from GenBank, were also
included (Table 3).
Results
Morphological variation
The first morphological analysis, based on identifying
461 hake employing tabulated characteristics used
in previous studies to separate putative species
M. angustimanus and M. productus, did not provide
conclusive identification at the species level for most
of the specimens (72%; 332 specimens). This uncer-
tainty results from overlapping morphological char-
acteristics for these species.
The second morphological analysis, based on
PCA, showed that the first two components explained
52.7% of the variance (CP1 =31.2%; CP2 =21.5%;
Fig. 2). In the positive region of the first component,
specimens from the northern locations (Washington,
Oregon, Eureka, and San Francisco), Bahı
´a Sebastia
´n
Vizcaı
´no, and the Upper Gulf of California were
found; these specimens were identified by the large
number of fin rays in the second dorsal and anal fins
(2D and A, respectively) and small values for head
length (HL/SL), length of the upper jaw (ML/SL),
pre-orbital length (POL/SL), inter-orbital distance
(IO/HL), and length from the tip of snout to pectoral
fin insertion and to anal fin origin (S-Pec/SL and S-A/
SL; characteristics contribute to a variance C0.6). In
the negative region of the first component, hake of
Baja California Sur and Costa Rica displayed inverse
values with respect to the axis. The positive region of
the second component separated most specimens of
the Upper Gulf of California from other locations.
Table 3 List of accession
numbers of homologous
sequences retrieved from
GenBank for each one of
the gene regions analyzed
(1) Unpublished data
(GenBank); (2) Teletchea
et al. (2006); (3) Campo
et al. (2007); (4)
Unpublished data
(GenBank); (5) Rosel and
Kocher (2002); and (6)
Unpublished data
(GenBank)
Species Cyt-bCOI 16S rDNA
M. merluccius DQ197962 (1) DQ174012 (2) DQ274031 (3)
M. merluccius DQ174063 (2) DQ174013 (2) DQ274032 (3)
M. merluccius DQ274033 (3)
M. merluccius DQ274034 (3)
M. paradoxus EF362909 (3) DQ274035 (3)
M. senegalensis EF362908 (3) DQ274040 (3)
M. capensis EF362907 (3) DQ274022 (3)
M. capensis DQ274023 (3)
M. albidus EF362893 (3) DQ274016 (3)
M. albidus DQ274017 (3)
M. albidus DQ274018 (3)
M. polli EF362915 (3) DQ274036 (3)
M. polli DQ274037 (3)
M. bilinearis EF362886 (3) DQ174009 (2) DQ274021 (3)
M. bilinearis DQ174059 (2) DQ174010 (2) AF420456 (5)
M. hubbsi EF362890 (3) DQ274026 (3)
M. hubbsi DQ274027 (3)
M. australis EF362883 (3) DQ274019 (3)
M. gayi EF362887 (3) DQ174011 (2) DQ274024 (3)
M. gayi DQ274025 (3)
M. productus DQ174064 (2) DQ174014 (2) DQ274039 (3)
M. productus DQ174065 (2) DQ174015 (2) EF458337 (6)
M. productus EF458338 (6)
M. angustimanus EF362897 (3) DQ274041 (3)
M. hernandezi AY821672 (4)
Rev Fish Biol Fisheries (2011) 21:259–282 265
123

These individuals were defined by large values of the
length of the pectoral fin and diameter of the orbit
(LPec/SL, DO/SL, and DO/HL, with eigenvalues
C0.6). The third component (not shown in Fig. 2),
representing only 10.2% of explained variance, did
not show characteristics with contributions with
eigenvalues C0.6 and did not demonstrate differences
among locations. PCA, despite a tendency to form
three groups: north (W, O, E, SF, and to a lesser
extent VIZ), south (CR and BCS), and eastern (UG),
showed broad overlap of morphological characteris-
tics with individuals belonging to localities, such as
SF, VIZ, and CR in the multidimensional spatial
distribution (Fig. 2). Other characteristics (1D, RPec,
LPec/HL, ML/HL, IO/DO, POL/HL, and GRI) con-
tributed little to the analysis of the total variation of
principal components (eigenvalues \0.6).
In the third analysis, the characteristics that
explained the largest percentage of variance (charac-
teristics with eigenvalues C0.6) were: 2D, A, HL/SL,
ML/SL, POL/SL, S-Pec/SL, S-A/SL, LPec/SL, DO/
SL, IO/HL, and DO/H. Graphic comparisons showed
significant differences (confidence intervals did not
overlap) between the characteristics from some
collection sites that were geographically farther away
(Fig. 3). However, when comparing two nearby
collection sites, the overlap in the variation intervals
of the characteristics was evident. The locality mean
shifted, showing a latitudinal cline with means
increasing from south to north for meristic charac-
teristics (2D and A). For morphometric characteris-
tics (HL/SL, ML/SL, POL/SL, S-Pec/SL, S-A/SL,
LPec/SL, DO/SL, IO/HL, and DO/HL), a reverse
trend occurred, whereby means decreased from south
to north (Fig. 3). This clinal effect was corroborated
with a regression analysis between morphometric and
meristic characteristics with respect to the latitude of
the collection sites, where most characteristics
showed a significant latitudinal cline, P\0.001
(2D: r=0.66, A: r=0.62; HL/SL: r=-0.67;
ML/SL: r=-0.6547; OLP/SL: r=-0.49; S-Pec/
SL: r=-0.57; S-A/SL: r=-0.47; LPec/SL: r=-
0.32; DO/SL: r=-0.34; IO/HL: r=-0.40). The
exception was DO/HL (r=0.05, P=0.25). Some
morphometric characteristics showed strong depen-
dence on body size (allometric effect), with a high
level of significance (P=0.0001; HL/SL: r=-
0.58; ML/SL: r=-0.48; OLP/SL: r=-0.43; S-
Pec/SL: r=-0.50; DO/SL: r=-0.73; IO/HL:
r=-0.34, except DO/HL (r=-0.47, P=0.25).
In general, smaller specimens had a proportionally
larger head length, orbital diameter, upper jaw, pre-
orbital length, and length from tip of snout to pectoral
fin insertion. The opposite occurred with larger
specimens. The width and length of the otolith
(sagittae), although not available for analysis at all
localities, also showed a strong association with size
(WO/SL r=0.82, P=0.0001; LO/SL r=0.70,
P=0.0001), proportionally larger in small hake than
large hake. This bias means that these characteristics
are not useful for separating hake species because they
vary with body size. They could be useful for
identifying different geographic populations.
Genetic variation
We analyzed 110 sequences for each mitochondrial
gene as well as the combined fragment. The number
of nucleotides was 351 bp for Cyt-b, 402 bp for COI,
449 bp for 16S, and 1202 bp for the combined
fragment (Cyt-b ?COI ?16S). All the nucleotide
positions were clearly aligned (there were no gaps).
The partition homogeneity test allowed clustering of
the three genes (P=0.93). Nucleotide saturation was
not detected in any of the three genes or the combined
-8
-6
-4
-2
0
2
4
6
8
-8-6-4-202468
PC2 (21.5%)
PC1 (31.2%)
CR BCS VIZ UG SF E O W
Fig. 2 Principal components analysis based on morphological
data. Individuals are projected onto the plane formed by the
first two principal components axes. Washington (W(open
square)), Oregon (O(filled square)), Eureka, California (E
(open circle)), San Francisco, California (SF (filled circle)),
Upper Gulf of California (UG (plus)), Bahı
´a Sebastian
Vizcaino, Baja California (VIZ (open triangle)), southern part
of the State of Baja California Sur state (BCS (minus)), and
Costa Rica (CR (filled diamond))
266 Rev Fish Biol Fisheries (2011) 21:259–282
123

fragment (results not shown). Nucleotide diversity (p)
and haplotype diversity (h) were 0.004 and 0.78 for
Cyt-b; 0.003 and 0.62 for COI; 0.001 and 0.43 for
16S; and 0.002 and 0.89 for the combined fragment.
A large number of mitochondrial lineages or haplo-
types were present for the three genes and the
combined fragment (22, 19, 16, and 48 for the Cyt-b,
COI, 16S, and combined fragment, respectively;
Tables 4,5). However, haplotypes differed by one
to three mutational steps and almost all substitutions
were transitions (C-T or A-G); only one transversion
was observed in each gene. Each gene and the
combined fragment showed a common haplotype
(H1) with high frequency (42.7, 60, 74.5, and 30.9%
for Cyt-b, COI, 16S, and combined fragment,
respectively). This common haplotype was present
in all samples, except for COI in Costa Rica and the
combined fragment in Costa Rica and Upper Gulf of
California. Shared haplotypes were also observed
(low frequencies) between different sites, although
there were a large number of unique haplotypes in the
northern and southern areas. For example, Cyt-b
haplotypes H7–H11 were found in CR, BCS, and
VIZ; haplotypes H12–H18 were unique to UG, and
haplotypes H19–H22 occurred in SF, E, O, and W.
For all genes studied, the Upper Gulf of California
had the largest number of unique haplotypes, the
most frequent (H12 occurred in 7 of 23 specimens for
Cyt-b; H10 occurred in 13 of 23 for COI; H7
occurred in 12 of 23 for 16S; and H20 occurred in 3
of 23 for the combined fragment).
The genetic distances estimated with the Tamura-
Nei model at the inter-specific level for hakes of the
genus Merluccius (M. merluccius,M. senegalensis,
M. capensis,M. polli,M. paradoxus,M. bilinearis,
M. albidus,M. australis,M. hubbsi,M. gayi and
M. productus) are shown in Table 6; their mean and
range values were 9% (1.2–14%), 7.8% (2.0–11.3%),
2.4% (0.6–4.3%), and 5.8% (1.4–8.1%) for Cyt-b,
COI, 16S, and the combined fragment, respectively.
The genetic distances at the intra-specific level for
GenBank sequences of M. merluccius,M. capensis,
CR
BCS
VIZ
UG
SF
E
O
W
30 32 34 36 38 40 42 44
2D
32 34 36 38 40 42 44 46
A
24 26 28 30 32 34 36
HL/SL
10 11 12 13 14 15 16 17 18 19
ML/SL
7891011121314
POL/SL
CR
BCS
VIZ
UG
SF
E
O
W
22 24 26 28 30 32 34 36 38
S-Pec/SL
36 40 44 48 52 56 60
S-A/SL
14 16 18 20 22 24 26 28 30
LPec/SL
345678
DO/SL
16 18 20 22 24 26 28 30 32 34
IO/HL
Fig. 3 Comparisons of characteristics that explain the highest
percentage of variance in PCA (characteristics having eigen-
values C0.6). The central dots are means, the box represents
the 95% confidence intervals, and the whisker is the range for
each collection site. Collection sites are coded in Fig. 1.
Meristic characteristics: 2D is the number of second dorsal fin
rays and A is the number of anal fin rays. Morphometric
characteristics expressed as percentages of standard length
(SL) and head length (HL): HL/SL is head length, ML/SL is
length of the upper jaw, POL/SL is pre-orbital length, S-Pec/SL
is the length from tip of snout to pectoral fin insertion, S-A/SL
is length from tip of the snout to anal fin origin, LPec/SL is
length of pectoral fin, DO/SL is diameter of orbit, and IO/HL is
inter-orbital distance
Rev Fish Biol Fisheries (2011) 21:259–282 267
123

Table 4 Variable sites and haplotype frequencies of three mitochondrial DNA genes of hake from the northeastern Pacific
A
Cyt-bVariable sites Sample sites
Haplotype 16 30 72 78 84 145 168 189 198 207 212 264 294 295 300 331 336 CR BCS VIZ UG SF E O W Total
H1 G C G C C G T A G T A C G T G C G 4 10 2 1 10 3 10 7 47
H2 –––––– – – – – – – – – – – A 3 1 1 5
H3 –––––– – – A – – – – – – – – 2 5 41 2418
H4 –T––––––––––––––– 1 1 215
H5 –––––– – – A – – – – – – – A 2 1 3 6
H6 –––T––––––––––––– 1 1 1 3
H7 –––––– – – A – – – – – A – – 1 1
H8 ––––T– – – – – – T – – – – – 1 1
H9 –––––– – – – C – – – – – – – 1 1
H10 –––––– – G – – – – – – – – – 1 1
H11 –––––– – – A – – – – – – G – 1 1
H12 ––A–––––A–––––––A 7 7
H13 ––A–––––A–––––––– 4 4
H14 –––––– C – A – – – – – – – – 1 1
H15 –TA–––––A–––––––– 1 1
H16 ––A––– – – A – G – – – – – – 1 1
H17 –––––– – – – – – – A – – – – 1 1
H18 ––A–––––A–––––A–– 1 1
H19 –––––A – – – – – – – – – – – 2 2
H20 –––T–––––––––C––– 1 1
H21 A––––– – – – – – – – – – – – 1 1
H22 ––A–––––––––––––– 11
B
COI Variable sites Sample sites
Haplotype 27 46 48 72 93 117 120 144 150 174 198 262 321 342 348 360 363 369 CR BCS VIZ UG SF E O W Total
H1 CCATTGTGCAGGGCCCAT 7 7 3 174121666
H2 ––G–––––––––––––––43 1 8
H3 –TG–––––––––––––––21 1 4
268 Rev Fish Biol Fisheries (2011) 21:259–282
123

Table 4 continued
B
COI Variable sites Sample sites
Haplotype 27 46 48 72 93 117 120 144 150 174 198 262 321 342 348 360 363 369 CR BCS VIZ UG SF E O W Total
H4 –––C–– – ––––––––––– 1 1 2 4
H5 ––G–––––T–––––––––1 1
H6 ––G––A–––––––––T––1 1
H7 ––G–––––––––A––––– 1 1
H8 ––G–––––––––––T––– 1 1
H9 –TG–––C–––––––––G– 1 1
H10 T–G––––––––––––––– 13 13
H11 T–G–––––––A––––––– 2 2
H12 T–G–C––––––––––––– 1 1
H13 T–G––––A–––––––––– 1 1
H14 T–G–––––––––A––––– 1 1
H15 T–G––––––––––––––C 1 1
H16 T–G–––––––T––––––– 1 1
H17 –––––– – ––––A–––––– 1 1
H18 –––––– – – – G – – – ––––– 1 1
H19 –––––– – ––––––T–––– 11
C
16S rDNA Variable sites Sample sites
Haplotype 146 150 155 172 181 196 229 240 257 258 273 296 315 385 435 CR BCS VIZ UG SF EUR O W Total
H1 TGCGGAGCCGAGT T G7 138 2174 151682
H2 ––––––A–––––––– 112
H3 –––––––––A–––––1 1
H4 ––––A–––––––––– 1 1
H5 –––––––––––A––– 1 1
H6 –––––––T––––––A 1 1
H7 –––A––––––G–––– 12 12
H8 –––A––––––––––– 2 2
H9 ––––––––T–––––– 1 1
Rev Fish Biol Fisheries (2011) 21:259–282 269
123

M. polli,M. bilinearis,M. albidus,M. hubbsi,
M. gayi, and M. productus had values under 1.0%
for sequence divergence; for example, mean dis-
tances and ranges revealed for Cyt-b, COI, 16S, and
the combined fragment values of 0.6% (0.4–0.8%),
0.4% (0.3–0.6%), 0.3% (0.2–0.4%), and 0.3% (0.09–
0.5%), respectively. Similar small means and ranges
occurred in all mitochondrial haplotypes that we
examined, with 0.7% (0.3–1.1%) for Cyt-b; 0.7%
(0.2–1.4%) for COI; 0.5% (0.2–1.3%) for 16S; and
0.4% (0.09–0.9%) for the combined fragment. Like-
wise, when genetic distances were compared between
GenBank sequences of M. productus,M. angustim-
anus and M. hernandezi, small values were obtained,
similar to those observed at the intra-specific level,
with a mean and range of distance of 0.5% (0.4–
0.8%) for Cyt-band 0.4% for 16S.
The distribution of genetic distances in the three
comparison groups (Haplotype, Within Species, and
Between Species) for the three genes and the
combined fragment did not exhibit a normal distri-
bution; hence, non-parametric analysis was used to
assess significant differences (Fig. 4). The variation
of the distances were significant between the three
groups using the Kruskal-Wallis non-parametric
statistical test for ANOVA by rank and median tests:
Cyt-b(H=172.7, df =2, n=309, P=0.000);
COI (H=20.2, df =2, n=179, P=0.0000); 16S
(H=120.2, df =2, n=193, P=0.000); and com-
bined fragment (H=18.7, df =2, N=1136,
P=0.0001). However, among Haplotype and
Within-Species groups, no significant differences
were found for Cyt-b, COI, and the combined
fragment (Fig. 4), except for the 16S gene
(P=0.04). In contrast, the Between-Species group
was significantly different for the Haplotype and
Within-Species groups (P\0.01) for Cyt-b, COI,
16S, and the combined fragment (Fig. 4). Of partic-
ular interest is that distances among M. productus,
M. angustimanus, and M. hernandezi were not
included in the Between-Species group comparisons
because their genetic distance is smaller than 0.8%. If
they were included, the range would increase the
Between-Species group and would cause overlap
with the Within-Species group (results not shown).
Topologies for the phylogenetic trees obtained
from the combined fragment (Cyt-b ?COI ?16S)
and reconstructed under the NJ and Bayesian criteria
were similar (Fig. 5shows only the Bayesian tree
Table 4 continued
C
16S rDNA Variable sites Sample sites
Haplotype 146 150 155 172 181 196 229 240 257 258 273 296 315 385 435 CR BCS VIZ UG SF EUR O W Total
H10 –––A––––––G––C– 1 1
H11 –––A–GA–––G–––– 1 1
H12 –––A––––T–G–––– 1 1
H13 ––TA––––––G–––– 1 1
H14 –A–A––––––G–––– 1 1
H15 –––A––––––G–A–– 1 1
H16 C–––––––––––––– 1 1
Sampling locations codes as in Table 2. (A) Cyt-b; (B) COI; and (C) 16S rDNA
270 Rev Fish Biol Fisheries (2011) 21:259–282
123

Table 5 The variable sites and haplotype frequencies of combined fragment (Cyt-b?COI ?16S rDNA) of hake from northeastern
Pacific
Cyt-b?COI ?
16S rDNA
Variable sites
Haplotype 16 30 72 78 84 145 168 189 198 207 212 264 294 295 300 331 336 345 346 377 396
H1 GCGCCGTAGTACGT GCGAACC
H2 –––––– – ––––––––––––––
H3 –––––– – –––––––––––––T
H4 –––––– – –––––––––A––––
H5 –T–––– – ––––––––––––––
H6 –––––– – – A – – ––––––––––
H7 –––––– – – A – – –––––A––––
H8 –––T–––––––––––––––––
H9 –––––– – ––––––––A ––––
H10 –––––– – – A – – –––––––––T
H11 –––––– – ––––––––––––––
H12 –––––– – ––––––––––––––
H13 –––––– – ––––––––––––––
H14 ––––T––––––T––T––TT––
H15 –––––– – –––––––––––––T
H16 –––––– – –––––––––A––––
H17 –––––– – G – – – –––– T
H18 –––––– – – A – – –––– A –
H19 –––––– – – A – – ––––G––– –
H20 ––A–––––A––––––– ––T–
H21 ––A–––––A–––––––A––T–
H22 –––––– – – A – – ––––––––T–
H23 –––––– C – A – – ––––––––T–
H24 –––––– – ––––––––––––––
H25 –––––– – – A – – ––––––––T–
H26 ––A–––––A–––––––A––T–
H27 –––––– – – A – – ––––––––T–
H28 ––A–––––A–––––––A––T–
H29 –––––– – – A – – –––––A––T–
H30 –––––– – – A – – ––––––––T–
H31 ––A–––––A–––––––A––T–
H32 ––A–––––A–––––––A––T–
H33 ––A–––––A–––––––A––T–
H34 –TA–––––A––––––––––T–
H35 ––A–––––A–G––––––––T–
H36 –––––– – – – – – – A A – –––––
H37 –––––– – – A – – ––––––––T–
H38 ––A–––––A––––––––––––
H39 ––A–––––A–––––A––––T–
H40 –––––A – – – – ––––––––––
H41 –––––– – – A – – ––––––––––
H42 –––––– – – – – ––––––––––
Rev Fish Biol Fisheries (2011) 21:259–282 271
123

Table 5 continued
Cyt-b?COI ?
16S rDNA
Variable sites
Haplotype 16 30 72 78 84 145 168 189 198 207 212 264 294 295 300 331 336 345 346 377 396
H43 –––––– – – – – ––––––––––
H44 –––––– – – – – ––––––––––
H45 –––––– – – – – –––C––––––
H46 A––––– – – – – ––––––––––
H47 –T–––– – – – – ––––––––––
H48 ––A––––– ––––––––––––
Cyt-b?COI ?
16S rDNA
Variable sites
Haplotype 398 422 443 467 470 494 500 524 548 612 671 692 698 710 713 719 796 898 902 907 924
H1 ATT GTGCAGGGCCCATGTGCG
H2 G––––––––––––––––––––
H3 G––––––––––––––––––––
H4 G––––––––––––––––––––
H5 –––––––––––––––––––––
H6 –––––––––––––––––––––
H7 –––––––––––––––––––––
H8 –C–––––––––––––––––––
H9 G–––––T––T–––––––––––
H10 G––––––––––––––––––––
H11 G––A–––––––––––––––––
H12 G––––––––––––A–––––––
H13 G––––––––––––––––––––
H14 G––––––––––––––T–––––
H15 G–––C––––––––––––––––
H16 –––––––––––––––––––––
H17 G––––––––––––––––––––
H18 ––––––––––––––––G––––
H19 –––––––––––––––––––––
H20 G–––––––––––––––––––A
H21 G–––––––A–––––––––––A
H22 G–––––––––––––––––––A
H23 G–––––––––––––––––––A
H24 –––––––––––––––––––––
H25 G–C–––––––––––––––––A
H26 G––––A––––––––––––––A
H27 G–––––––A––––A––––––A
H28 G–––––––––––––––––––A
H29 G–––––––––––––––––––A
H30 G–––––––––––––––––––A
H31 G–––––––––––––––––––A
H32 G–––––––––––––––––––A
H33 G––––––––––––––––––TA
272 Rev Fish Biol Fisheries (2011) 21:259–282
123

Table 5 continued
Cyt-b?COI ?
16S rDNA
Variable sites
Haplotype 398 422 443 467 470 494 500 524 548 612 671 692 698 710 713 719 796 898 902 907 924
H34 G–––––––––––––––––A–A
H35 G–––––––––––––––––––A
H36 ––––––––––––––––––––
H37 G–––––––––––––––––––A
H38 –––––––––––––––––––––
H39 G––––––––––T––––––––A
H40 –––––––––––––––––––––
H41 –––––––––––––––––C–––
H42 –––––––––––––––––––––
H43 –––––––––A–––––––––––
H44 –––––––G–––––––––––––
H45 –C–––––––––––––––––––
H46 –––––––––––––––––––––
H47 ––––––––––––––T––––––
H48 –––––––––––––––––––––
Cyt-b?COI ?
16S rDNA
Variable sites Sample sites
Haplotype 933 948 981 992 1009 1010 1025 1048 1067 1137 1187 CR BCS VIZ UG SF E O W Total
H1 G A G C C G A G T T G 5 2 10 3 7 7 34
H2 ––––– – – – – – – 12 3
H3 ––––– – – – – – – 11 2
H4 ––––– – – – – – – 3 1 4
H5 ––––– – – – – – – 1 1 2 4
H6 ––––– – – – – – – 2 311512
H7 ––––– – – – – – – 1 34
H8 ––––– – – – – – – 1 1 1 3
H9 ––––– – – – – – – 1 1
H10 ––––– A – – – – – 1 1
H11 ––––– – – – – – – 1 1
H12 ––––– – – – – – – 1 1
H13 A–––– – – – – – – 1 1
H14 ––––– – – – – – – 1 1
H15 ––––– – – – – – – 1 1
H16 ––––– – – – – – – 1 1
H17 ––––– – – A – – 1 1
H18 –––T––––––A 1 1
H19 ––––– – – – – – – 1 1
H20 ––––– – G – – – – 3 3
H21 ––––– – G – – – – 2 2
H22 ––––– – G – – – – 1 1
H23 ––––– – – – – – – 1 1
H24 ––––T – – – – – – 1 1
Rev Fish Biol Fisheries (2011) 21:259–282 273
123

analysis). The tree bifurcated into two main clades
with weak bootstrap support (58%). Nucleotide
divergence between these two clusters averaged
0.6% (obtained by Tamura-Nei genetic distance);
this difference was the result of only three nucleotide
substitutions. The biggest clade, with a bootstrap
support of 88%, included all haplotypes found from
Washington to Costa Rica (H1–H19 and H40–48; see
haplotypes in Table 5), three unique haplotypes from
the Upper Gulf (H24, H36, and H38), and two for M.
productus from GenBank (Mpro1 and 2). The second
clade had smaller bootstrap support (54%) and
contained 15 unique haplotypes from the Upper Gulf
(H20–H22, H26–H35, H37, and H39). Two unique
haplotypes from the Upper Gulf were unresolved
(H23 and H25). The difference between these two
haplotypes relative to the rest was one mutational
step (transition C-T) at different sites. Overall, all the
Upper Gulf haplotypes had three mutational steps that
differed from the rest of the haplotypes, although they
share one to three mutation steps, similar to other
haplotypes (see Table 5).
Discussion
Morphological evidence
Until now, species differentiation for northeastern
Pacific hake has been based on morpholgical studies
that used the average values of a combination of
morphological characteristics, such as: the number of
anal fin rays of the first and second dorsal; gill rakers
number; the number of vertebrae; and the head length
Table 5 continued
Cyt-b?COI ?
16S rDNA
Variable sites Sample sites
Haplotype 933 948 981 992 1009 1010 1025 1048 1067 1137 1187 CR BCS VIZ UG SF E O W Total
H25 – – – – – – – – – – 1 1
H26 – – – – – G – – – – 1 1
H27 – – – – – G – – – – 1 1
H28 –––– –G––C– 1 1
H29 – – – – – G – – – – 1 1
H30 G A – – G – – – – 1 1
H31 ––––T – G – – – – 1 1
H32 ––––– – G – – – – 1 1
H33 ––––– – G – – – – 1 1
H34 ––––– – G – – – – 1 1
H35 ––––– – G – – – – 1 1
H36 ––––– – – – – – – 1 1
H37 ––––– – G – – – – 1 1
H38 ––––– – – – – – – 1 1
H39 ––––– – G – A – – 1 1
H40 ––––– – – – – – – 2 2
H41 ––––– – – – – – – 1 1
H42 ––A–––––––– 1 1
H43 ––––– – – – – – – 1 1
H44 ––––– – – – – – – 1 1
H45 ––––– – – – – – – 1 1
H46 ––––– – – – – – – 11
H47 ––A–––––––– 11
H48 ––––– – – – – – – 11
Sampling locations codes as in Table 2
274 Rev Fish Biol Fisheries (2011) 21:259–282
123

as a proportion of the standard length (Ayres 1855;
Garman 1899; Ginsburg 1954; Vrooman and Paloma
1977; Inada 1981,1995; Mathews 1985; Lloris et al.
2003; Balart-Pa
´ez 2005). Our analyses, using the
above characteristics failed to distinguish species of
hake in the northeastern Pacific since most charac-
teristics overlapped. Additionally, previous studies
were limited to the analysis of small sample sizes and
the use of specimens from isolated localities, as was
the case in the original descriptions of M. productus
(Ayres 1855) and M. angustimanus (Garman 1899).
For the latter species, no comparisons with other
species have been made. Ginsburg (1954) expanded
the geographical coverage, but his sample size was
small, and although later studies (i.e., Vrooman and
Paloma 1977; Inada 1981; Mathews 1985; Balart-
Pa
´ez 2005) sampled a larger number of specimens,
these specimens were limited to the coast of the Baja
California Peninsula (M. angustimanus) or from
California to Oregon (M. productus; Table 1). Our
study attempted to overcome these shortcomings by
analyzing more specimens from a larger part of the
nominal range of distribution. This larger sampling
led to the conclusion that the morphological
Table 6 Matrix of Tamura-Nei’s genetic distances of Merluccius species (in %)
A
Cytb\16S Mmer
(4)
Msen Mcap
(2)
Mpol
(2)
Mpar Mbil
(2)
Malb
(3)
Maus Mhub
(2)
Mgay
(2)
Mang Mher Mpro
(3)
H
(16)
Mmer
(2)
[0.4\0.4] 1.1 1.3 2.2 2.1 4.0 3.2 2.8 2.2 2.8 3.1 – 3.0 3.1
Msen 1.4 – 1.3 2.2 2.1 4.3 3.7 3.5 2.4 3.0 3.5 – 3.4 3.5
Mcap 4.8 5.2 [–\0.2] 1.4 1.7 3.7 3.1 2.7 1.8 2.4 2.9 – 2.8 2.9
Mpol 7.6 7.4 7.8 [–\0.2] 0.6 4.0 3.7 3.1 2.4 2.7 3.1 – 3.0 3.1
Mpar 7.6 8.3 7.3 4.0 – 4.1 4.2 3.0 2.7 2.8 3.0 – 2.9 3.0
Mbil (2) 12.5 12.2 13.1 10.8 10.8 [0.8\0.2] 3.1 2.4 2.3 1.8 2.0 – 1.9 2.0
Malb 13.0 13.6 11.7 10.5 11.0 7.3 [–\0.3] 3.0 2.1 2.7 2.5 – 2.4 2.5
Maus 13.2 13.0 13.3 11.0 10.9 6.9 4.4 – 1.7 0.8 0.9 – 0.8 0.9
Mhub 12.8 12.6 12.9 10.5 11.8 7.3 4.0 4.8 [–\0.2] 0.9 1.7 – 1.6 1.8
Mgay 12.9 12.7 13.0 10.6 10.5 6.9 6.5 4.8 5.6 [–\0.2] 0.8 – 0.7 0.9
13.9 13.6 13.9 11.1 11.5 6.9 5.6 4.8 4.8 1.6 – – 0.4 0.3
Mang
Mher
13.4 13.2 13.5 10.7 11.0 6.5 5.2 4.4 4.4 1.2 0.4 – – –
Mpro (3) 13.9 13.7 14.0 11.1 11.3 6.9 5.6 4.5 4.2 1.6 0.8 0.4 [0.5\0.3] 0.5
H (22) 14.1 13.8 14.1 11.3 11.4 7.2 6.0 4.7 4.4 1.9 1.1 0.8 0.5 [0.7\0.5]
B
COI\(Cytb ?COI ?16S) Mmer (2) Mbil Mgay Mpro (2) H (48)
Mmer (2) [0.6\0.5] 8.1 8.1 7.9 7.9
Mbil (2) 9.1 [0.3\–] 5.1 4.8 4.9
Mgay 11.3 7.9 – 1.4 1.4
Mpro 9.9 6.6 2.0 [–\0.09] 0.3
H (19) 10.1 6.7 2.1 0.5 [0.7\0.4]
Mmer =Merluccius merluccius; Msen =M. senegalensis; Mcap =M. capensis; Mpol =M. polli; Mpar =M. paradoxus;
Mbil =M. bilinearis; Malb =M. albidus; Maus =M. Australis; Mhub =M. hubbsi; Mgay =M. gayi; Mang =M. angustimanus;
Mher =M. hernandezi; Mpro =M. productus; and H =haplotypes determined in this study. Numbers in parenthesis are the
number of haplotypes within species. (A) Above diagonal: genetic distance of 16SrDNA sequences; below diagonal: distances of
Cyt-bsequences; and in diagonal: genetic distance within species [Cytb\16SrDNA sequences]. (B) Above diagonal: genetic distances
of combined sequences (Cyt-b?COI ?16SrDNA); below diagonal: genetic distances of COI sequences; and in diagonal: genetic
distances within species [COI\Cyt-b?COI ?16SrDNA]
Rev Fish Biol Fisheries (2011) 21:259–282 275
123

characteristics traditionally used as species discrim-
inators exhibited a shift from north to south in mean
values and considerable overlap in ranges (Fig. 3).
We observed that this continuous overlap of charac-
teristics does not provide sufficient evidence to
robustly identify species along this latitudinal gradi-
ent. Our additional morphological analyses (PCA,
graphics method, and regression) confirmed the
latitudinal trend (cline) of the characteristics.
The most common method used by taxonomists to
assign specimens to taxa is by comparison of
anatomical, meristic, and morphometric characteris-
tics. However, morphometric and meristic character-
istics are often difficult to interpret because they
show great variability, a high degree of overlap
between species, and are subject to complex, non-
genetic sources of variation because they are strongly
influenced by biotic factors (Swain 1992; Lankford
et al. 2001; Yamahira et al. 2006) and abiotic factors
(Franca 1962; Hadfield et al. 1979; Jordan et al. 2003;
Lloris et al. 2003; McDowall 2008). Therefore,
caution must be taken with their use for taxonomic
purposes, especially if there are limited studies that
have examined their allometric and clinal variation.
Our PCA analyses did not support two separate
species (M. productus and M. angustimanus). Instead,
it supported a single species with three contiguous
and overlapping groups in three geographical areas: a
northern group (W, O, E, SF, and to a lesser extent
VIZ), a southern group (BCS and CR) and an eastern
group (UG and a few from VIZ). VIZ hake indicate
that, in this region, there may be a mix of northern
and Upper Gulf groups. The northern hake exhibited
larger meristic values (2D and A) and smaller
morphometric values (HL/SL, ML/SL, POL/SL, IO/
HL, S-Pec/SL, and S-A/SL), while southern hake
were morphologically the opposite. Upper Gulf hake
mainly showed larger values for LPec/SL, DO/SL
and DO/HL. These results suggest that these charac-
teristics differentiate regional intra-specific morpho-
types, probably in response to ocean-climate
variation within the distributional range. Morpholog-
ical variation of similar groups within a species has
been noted in other studies by measuring the means
and ranges along a geographical gradient, as in
Chilean hake, anchovies, and sardines (Hubbs 1925;
Hubbs and Hubbs 1953; Blackburn 1967; Martı
´nez
1976; Hadfield et al. 1979).
In the northeastern Pacific, some fish undertake
large latitudinal migrations, such as anchovy and
sardines. These show a clear population structure
throughout their distribution with a latitudinal variation
in their meristic and morphometric characteristics
(Hubbs 1925; McHugh 1951; Miller 1956;Fe
´lix-Uraga
Haplotype
Within Species
Between Species
0
2
4
6
8
10
12
14
16
Cyt-
b
Tamura-Nei distance (%)
Median 25%-75% Min-Max
Haplotype
Within Species
Between Species
0
2
4
6
8
10
12
COI Tamura-Nei dIstance (%)
Haplotype
Within Species
Between Species
0
1
2
3
4
5
6
7
8
9
Cyt-
b
+COI+16S rDNA Tamura-Nei
distance (%)
Haplotype
Within Species
Between Species
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
16S rDNA Tamura-Nei
distance (%)
Fig. 4 Variation detected
in the Tamura-Nei distances
for the nucleotide sequences
of the analyzed regions in
Cyt-b, COI, 16S rDNA, and
Cyt-b?COI ?16S rDNA
in Merluccius species.
Groups of comparison:
haplotype (distances
between the haplotypes
detected in this study);
within species (distances of
the GenBank sequences
within each of the species);
and between species
(distances between the
GenBank sequences for the
different species). The small
black rectangles indicate
the median, the large
rectangles indicate quartiles
(25–75%), and whiskers
indicate the ranges
276 Rev Fish Biol Fisheries (2011) 21:259–282
123

et al. 2005). A similar situation could be operating to
account for the variation of hake in the northeastern
Pacific, since they also show latitudinal migration
patterns (Bailey et al. 1982; Saunders et al. 1997).
In our study, all meristic characteristics were
measured except the number of vertebrae. In other
studies, the data for this characteristic in M. produc-
tus and M. angustimanus also overlapped (Vrooman
and Paloma 1977; Inada 1981; Mathews 1985). This
characteristic, which has been used in several species
and is recognized as being strongly influenced by
many genetic and environmental factors, result in
complex patterns of variation (McDowall 2008) that
make discrimination between the hereditary and
environmental causes of variation difficult (Yamahira
et al. 2006).
Genetic evidence
The sequence analyses indicated high levels of
haplotype diversity, but low nucleotide diversity.
There were one or two haplotypes with high
frequency and many rare haplotypes in each of the
three genes and the combined fragment datasets. This
pattern of variation in mitochondrial DNA occurs at
the population level because there is rapid population
growth (expansion) after a period of low effective
population size, which enhances retention of new
mutations (Avise et al. 1984; Rogers and Harpending
1992). Examples are typically drawn from large
populations or entire species that contain one or two
prevalent haplotypes embedded in a cluster of small
branches that are one or a few mutations removed
from the common haplotypes (Grant and Bowen
1998). In a review of mtDNA diversity, this pattern is
a recurring pattern in widely distributed marine fishes
(Shields and Gust 1995). Alternative explanations
include: a large variance in reproductive success that
leads to the propagation of only a few haplotypes
(Shields and Gust 1995); overfishing (Camper et al.
1993); physical conditions in the pelagic realm
(Graves 1995); population bottlenecks (Gold et al.
Fig. 5 Bayesian tree
analysis based on the
combined fragment
haplotypes with
three mitochondrial genes
(Cyt-b?COI ?
16SrDNA; 1202 pb); M bil
is M. bilinearis; M gay is
M. gayi; and M pro 1 and 2
are M. productus combined
haplotypes retrieved from
GenBank; H1–H48 are
haplotypes of the combined
fragments from this study
and corresponds to
haplotypes listed in Table 5.
Bootstrap support values
that are C50% are shown
for each branch. Sampling
location codes are described
in Fig. 1.Small black
squares indicate unique
haplotypes in the Upper
Gulf of California
Rev Fish Biol Fisheries (2011) 21:259–282 277
123

1994); and other demographic events (Dodson et al.
1991). Northeastern Pacific hake have a wide
geographical distribution (Bailey et al. 1982; Saun-
ders et al. 1997) and variations of mitochondrial
DNA sequences that is similar to this pattern. This
suggests explanations about population history for the
species. Consequently, analysis of mtDNA sequence
diversity supports intra-specific variation of the
northeastern Pacific hake rather than inter-specific
variation. The large number of unique haplotypes in
the Upper Gulf of California likely occurred from
rapid population growth after a period of low
effective population size with retention of new
mutations; this population may be isolated by phys-
ical or environmental conditions. High polymorphism
and mixing of some of these haplotypes encourages
further analysis at the population level to explain this
pattern.
Genetic distances within Merluccius have been
estimated using different types of genetic markers.
With mtDNA markers, such as the control region, the
reported genetic distances (Tamura-Nei) within spe-
cies (intra-specific variation) are in the order of 0.3–
0.8% and 2–20% between hake species (inter-specific
variation); however, for M. merluccius, higher intra-
specific distances (0.5–1.8%) have been reported
(Quinteiro et al. 2000). For Cyt-b, Campo et al.
(2007) reported genetic divergence (Kimura 2-param-
eter) between species of hake range from 0.5 to
13.5%, the lowest value corresponding to the dis-
tances between M. angustimanus and M. productus
(sister species); however, sampling locations for the
sequences of M. angustimanus were not given.
The analyses combined new data with GenBank
data, providing a more detailed analysis of genetic
distance at taxonomic levels within Merluccius. The
results indicated a statistically significant differenti-
ation in the genetic distances between the groups
Within Species and Between Species, but no differ-
entiation between the Within-Species group and the
Haplotype group. The distances revealed a pattern of
increased nucleotide diversity between Within-Spe-
cies group and Between-Species group. Using genetic
distances for assessing taxonomic levels was
employed in other studies. Kartavtsev and Lee
(2006) detected a statistically significant variation
between mean intra-specific distances (1.55 ±0.56%
for Cyt-band 0.55 ±0.19% for COI) and mean inter-
specific distances (10.69 ±1.34% for Cyt-band
9.96 ±0.72% for COI) in vertebrate and invertebrate
species; Kartavtsev et al. (2007) reported that catfish
average inter-specific differences of 5.28 ±1.74.
Ward et al. (2005) reported that Australian fish
species averaged intra-specific genetic distance (Kim-
ura 2 parameters) for COI of 0.39% and inter-specific
distance of 9.93%. We studied differential rates of
divergence in 16S, Cyt-b, and COI and found that the
16S gene displayed the lowest rate of mutation; still,
we detect intra-specific and inter-specific variation in
all three markers, similar to the report by Peregrino-
Uriarte et al. (2007).
Phylogenetic analysis of the combined haplotypes
with combined haplotypes for M. bilinearis,M. gayi
and M. productus from GenBank resolved a bifur-
cated tree, showing two main clades and low
bootstrap support (58%). While this could be ascribed
to of two species, but to define a separation between
M. productus and M. angustimanus, we would expect
to find unique haplotypes from Costa Rica separated
from the other haplotypes, since there is a high
probability that M. angustimanus is present in Costa
Rica, as mentioned in the taxonomic literature.
However, this was not the case. Instead, we found
an Upper Gulf clade that deserved special mention
because it has unique haplotypes, which could
indicate a true specie or subspecies; however, some
other unique haplotypes from Upper Gulf hake were
present in the clade of Pacific hake and even more,
there were two additional unresolved haplotypes in
the tree (Fig. 5). We suggest that this is not a
monophyletic clade and neither should be considered
separate species, as defined by the phylogenetic
species concept (De Queiroz 2007). Rather, this tree
reflects two population units with some degree of
gene flow. One population inhabits offshore waters
from Washington to Costa Rica and the other inhabits
the upper Gulf of California. Previous studies of
M. productus genetic structure along the west coast of
the Unites States failed to find significant genetic
differentiation within the migratory Pacific stock;
instead, significant variations were found between
offshore and Puget Sound (inshore waters). This was
explained by a complex regional physiography that
provides an isolating mechanism for genetic differ-
entiation (Utter et al. 1970; Utter and Hodgins 1971;
Grant and Leslie 2001; Iwamoto et al. 2004). Overall,
it is common that Merluccius species show subdi-
vided populations around geographically complex
278 Rev Fish Biol Fisheries (2011) 21:259–282
123

coastlines (Rolda
´n et al. 1998; Lundy et al. 1999;
Grant and Leslie 2001; Cimmaruta et al. 2005), but
not along linear coastlines (Smith et al. 1979; Grant
et al. 1987). Consequently, the unique haplotypes in
the Upper Gulf does not indicate species-level
differentiation, but isolation caused by the Baja
California Peninsula and showing some gene flow,
as shown by shared haplotypes and haplotype tree
with low resolution. An alternative explanation is that
some offshore hake migrated into the Upper Gulf.
Using other molecular markers with higher mutation
rates, such as microsatellites, could resolve remaining
ambiguities.
An additional concern is whether to consider
subspecies designation for the Upper Gulf hake. Mayr
(1969) considered, that for a population to be
accepted as a valid subspecies, 75% of its population
should differ from the other geographically separate
populations of the same species. For Merluccius spp.,
this has been the criteria (Ginsburg 1954; Lloris et al.
2003). For example: M. merluccius merluccius
(Atlantic) and M. m. smiridus (Mediterranean) are
separated by the physiographic conditions of this
region; Merluccius australis australis (New Zealand)
and M. a. polylepis (southern coast of Chile); M. polli
cadenati and M. p. polli in the eastern Atlantic are
separated by 500 km; and M. gayi gayi and M. g.
peruanus in the southeastern Pacific are separated by
[1300 km. In the case of M. gayi subspecies, these
are accepted with reservations because their morpho-
logical differences are clinal in character and influ-
enced by environment conditions (Lloris et al. 2003).
This concern also applies to Atlantic and Mediterra-
nean hake, where genetic differentiation has been
attributed to environment features (Cimmaruta et al.
2005) or physical barriers to dispersal (Rolda
´n et al.
1998; Lundy et al. 1999). In the two studies, the
geographically separate populations are considered
stocks rather than subspecies.
For the northeastern Pacific area, clinal variation
in the morphological characteristics and intra-specific
genetic variation among haplotypes occur from
Washington to the southern part of the Baja Califor-
nia Peninsula, and include the upper Gulf of Califor-
nia and Costa Rica. We suggest that the currently
valid species (M. productus and M. angustimanus), as
well as M. hernandezi and dwarf hake are one
synonymous species. The close genetic relationships
of the haplotypes obtained in this study indicate a
high degree of gene flow and support the hypothesis
of a single hake species for this region. The proposal
of recognizing only a single species of hake was
suggested by F. H. Berry in 1965 (Inada 1981), for
the entire eastern Pacific. The proposal is that
M. productus,M. angustimanus,M. gayi gayi, and
M. g. peruanus are synonymous and should be
reclassified as a single species called M. gayi.This
hypothesis has not yet been tested and would require
a thorough analysis of all nominal species of hake,
including those extending the length of the entire east
Pacific. As a step to unraveling the taxonomy of hake,
our findings justify that the hake we describe in
the northeastern Pacific should be grouped under the
single name Merluccius productus according to the
priority principle (ICZN 1999) since it was first to
be described in the Northeastern Pacific (Ayres 1855).
We anticipate that a future reclassification would
require renaming all hake in the eastern Pacific,
probably as M. gayi (Guichenot 1848). Now that one
species can be recognized in the northeastern Pacific,
future work should be directed towards the genetic
analysis of population structure to determine the
magnitude of gene flow among Pacific populations.
Based on taxonomic and biological species con-
cepts (De Queiroz 2007), morphological and genetic
data do not support the hypothesis of two distinct
hake species in the northeastern Pacific (M. angu-
stimanus and M. productus) and suggests a single
taxonomic entity with a minor degree of morpholog-
ical and genetic intra-specific variation.
Acknowledgments Hake samples were provided by Manuel
Nevares, Mike Canino, Maria de L. Gonzalez-Rugge, Hugo
Cirilo, and Eva F. Isaak Vissuet. Special thanks to the crews of
the INP and CIBNOR vessels for cruises to obtain samples; the
Pacific Choice Seafood of California, and the commercial
shrimp fleets of the Pacific coast of Costa Rica. We thank
Lucı
´a Campos-Da
´vila at the Coleccio
´n Ictiolo
´gica at CIBNOR.
This project was funded by the Secretarı
´a de Agricultura,
Ganaderı
´a, Desarrollo Rural, Pesca y Alimentacio
´ndeMe
´xico
and the Consejo Nacional de Ciencia y Tecnologı
´a
(SAGARPA-CONACYT grant 2005-12058 to FJGDL. CASS
and MBC received CONACYT fellowships. Additional advice
was received from Gil Rosenthal, Darrin Hulsey, Ingo Schlupp,
and Miguel Cordoba.
References
Ahlstrom EH (1969) Distributional atlas of fish larvae in the
California Current region: jack mackerel, Trachurus
Rev Fish Biol Fisheries (2011) 21:259–282 279
123

symmetricus, and Pacific hake, Merluccius productus,
1951 through 1966. Calif Coop Ocean Fish Invest Atlas
11:v–xiii, 1–187
Ahlstrom EH, Counts RC (1955) Eggs and larvae of the Pacific
hake, Merluccius productus. Fish Bull 56:295–329
Aljanabi SM, Martinez I (1997) Universal and rapid salt-
extraction of high quality genomic DNA for PCR-based
techniques. Nucleic Acids Res 25:4692–4693
Ambrose DA (1996) Merlucciidae: hakes. In: Moser HG (ed)
The early stages of the fishes in the California Current
Region. Calif Coop Ocean Fish Invest Atlas 33, pp 508–511
Astrin JJ, Huber BA, Misof B, Klu
¨tsch CFC (2006) Molecular
taxonomy in pholcid spiders (Pholcidae, Araneae): eval-
uation of species identification methods using CO1 and
16S rRNA. Zool Scr 35:441–457
Avise JC, Neigel JE, Arnold J (1984) Demographic influences
on mitochondrial DNA lineage survivorship in animal
populations. J Mol Evol 20:99–105
Ayres WO (1855) Description of new species of Californian
fishes. Proc Calif Acad Nat Sci 1:23–77
Bailey KM, Francis RC, Stevens PR (1982) The life history
and fishery of Pacific whiting, Merluccius productus.
Calif Coop Ocean Fish Invest Rep 23:81–98
Balart-Pa
´ez EF (2005) Biologı
´a y ecologı
´a de la merluza ba-
jacaliforniana, Merluccius angustimanus Garman, 1899,
en la costa occidental de Baja California Sur, Me
´xico.
Doctoral Thesis, Universidad Auto
´noma de Nuevo Leo
´n,
Monterrey
Beal CH (1948) Reconnaissance of the geology and oil pos-
sibilities of Baja California, Mexico. Geol Soc Am Mem
31:1–138
Bermingham E, Lessios HA (1993) Rate variation of protein
and mitochondrial DNA evolution as revealed by sea
urchins separated by the Isthmus of Panama. Proc Natl
Acad Sci USA 90:2734–2738
Bermingham E, McCafferty SS, Martin AP (1997) Fish bio-
geography and molecular clocks: perspectives from the
Panamanian Isthmus. In: Kocher TD, Stepien CA (eds)
Molecular systematics of fishes. Academic Press, New
York, pp 113–128
Bernardi G, Findley L, Rocha-Olivares A (2003) Vicariance
and dispersal across Baja California in disjunct marine
fish populations. Evolution 57:1599–1609
Billington N, Hebert PDN (1991) Mitochondrial DNA diver-
sity in fishes and its implications for introductions. Can J
Fish Aquat Sci 48:80–94
Blackburn M (1967) Synopsis of biological information on the
Australian anchovy Engraulis australis (White). Calif
Coop Ocean Fish Invest Rep 11:34–43
Camper JD, Barber RC, Richardson LR, Gold JR (1993)
Mitochondrial DNA variation among red snapper (Lutj-
anus campechanus) from the Gulf of Mexico. Mol Mar
Biol Biotech 2:154–161
Campo D, Machado-Schiaffino G, Perez J, Garcia-Vazquez E
(2007) Phylogeny of the genus Merluccius based on
mitochondrial and nuclear genes. Gene 406:171–179
Cimmaruta R, Bondanelli P, Nascetti G (2005) Genetic struc-
ture and environmental heterogeneity in the European
hake (Merluccius merluccius). Mol Ecol 14:2577–2591
Cohen DM, Inada T, Iwamoto T, Scialabba N (1990) FAO
species catalogue. Vol. 10. Gadiform fishes of the world
(Order Gadiformes). An annotated and illustrated cata-
logue of cods, hakes, grenadiers and other gadiform fishes
known to date. FAO Fish Synop, Roma
Crisci JV, Lo
´pez-Armengol MF (1983) Introduccio
´n a la teorı
´a
y pra
´ctica de la taxonomı
´a nume
´rica. Secretarı
´a General
de la Organizacio
´n de los Estados Americanos. Programa
Regional de Desarrollo Cientı
´fico y Tecnolo
´gico, Wash-
ington, DC
De Queiroz K (2007) Species concepts and species delimita-
tion. Syst Biol 56:879–886
Dodson JJ, Carscadden JE, Bernatchez L, Colombani F (1991)
Relationship between spawning mode and phylogenetic
structure in mitochondrial DNA of north Atlantic capelin
Mallotus villosus. Mar Ecol Prog Ser 76:103–113
Fe
´lix-Uraga R, Quin
˜o
´nez-Vela
´zquez C, Hill KT, Go
´mez-Mu-
n
˜oz VM, Melo-Barrera FN, Garcı
´a-Franco W (2005)
Pacific sardine (Sardinops sagax) stock discrimination off
the west coast of Baja California and southern California
using otolith morphometry. Calif Coop Ocean Fish Invest
Rep 46:113–121
Felsenstein J (1985) Confidence limits on phylogenies: and
approach using the bootstrap. Evolution 39:783–791
Franca P (1962) Conside
´rations sur la taxonomie des Merluc-
cius de I’Atlantique oriental. Trabalhos do Centro de Bi-
ologia Piscatoria. Mem Junta Inv Ultram 2:7–48
Frost KJ (1981) Descriptive key to the otoliths of Gadid fishes
of the Bering, Chukchi, and Beaufort seas. Arctic 34:55–
59
Garman S (1899) Report on an exploration off the west coasts
of Me
´xico, Central and South America, and off the
Galapagos Islands, in charge of Alexander Agassiz, by the
U. S. Fish Commission steamer ‘‘Albatros’’ during 1891,
XXVI (The Fishes). Mem Mus Comp Zool Harvard Coll
24:1–431
Ginsburg I (1954) Whitings on the coasts of the America
continents. Fish Bull 96:187–208
Gold JR, Richardson LR, Furman C, Sun F (1994) Mitochon-
drial DNA diversity and population structure in marine
fish species from the Gulf of Mexico. Can J Fish Aquat
Sci 51:205–214
Grant WS, Bowen BW (1998) Shallow population histories in
deep evolutionary lineages of marine fishes: insights from
sardines and anchovies and lessons for conservation. J
Hered 89:415–426
Grant WS, Leslie RW (2001) Inter-ocean dispersal is an
important mechanism in the zoogeography of hakes
(Pisces: Merluccius spp). J Biogeogr 28:699–721
Grant WS, Leslie RW, Becker II (1987) Genetic stock structure
of the southern African hakes Merluccius capensis and M.
paradoxus. Mar Ecol Prog Ser 41:9–20
Graves JE (1995) Conservation genetics of fishes in the pelagic
marine realm. In: Avise JC, Hamrick JL (eds) Conserva-
tion genetics: case histories from nature. Chapman and
Hall, New York, pp 335–366
Guichenot A (1848) Historia fı
´sica y polı
´tica de Chile (en Gay,
C.). Zool 2:328–330
Hadfield AJ, Ivantsoff V, Johnston PG (1979) Clinal variation
in electrophoretic and morphological characters between
two nominal species of the genus Pseudomugil (Pisces:
Atheriniformes: Pseudomugilidae). Aust J Mar Freshwat
Res 30:375–386
280 Rev Fish Biol Fisheries (2011) 21:259–282
123

Ho J (1990) Phylogeny and biogeography of hakes (Merluc-
cius; Teleostei): a cladistic analysis. Fish Bull 88:95–104
Hubbs CL (1925) Racial and seasonal variation in the Pacific
herring, California sardine and California anchovy. Calif
Fish Game Comm Fish Bull 8:1–23
Hubbs CL, Hubbs C (1953) An improved graphical analysis
and comparison of series of samples. Syst Zool 2:49–57
Hubbs CL, Lagler KF (1958) Fishes of the Great Lakes Region.
Cranbrook Inst Sci Bull 26:1–213
Huelsenbeck JP, Ronquist R (2001) Mr Bayes: Bayesian
inferences of phylogeny. Bioinformatics 17:754–755
Huelsenbeck JP, Hillis DM, Jones R (1996) Parametric boot-
strapping in molecular phylogenetics: applications and
performance. In: Ferraris JD, Palumbi SR (eds) Molecular
zoology: advances, strategies and protocols. Wiley-Liss,
New York, pp 19–45
ICZN (1999) International code of zoological nomenclature. The
International Trust for Zoological Nomenclature, London
Inada T (1981) Studies on the Merluccid fishes. Bull Far Seas
Fish Res Lab 18:1–172
Inada T (1995) Merlucciidae. In: Fisher W, Krupp F, Schneider
W, Sommer C, Carpenter KE, Niem VH (eds) Guı
´a FAO
para la identificacio
´n de especies para los fines de la
pesca. Pacı
´fico centro-oriental. Volumen III. Vertebrados
Parte 2. FAO, Rome, pp 1272–1274
Iwamoto E, Ford MJ, Gustafson RG (2004) Genetic population
structure of Pacific hake, Merluccius productus, in the
Pacific Northwest. Environ Biol Fish 69:187–199
Johns GC, Avise JC (1998) A comparative summary of genetic
distances in the vertebrates from the mitochondrial cyto-
chrome bgene. Mol Biol Evol 15:1481–1490
Jordan AD, Moller PR, Nielsen JG (2003) Revision of the
Arctic cod genus Arctogadus. J Fish Biol 62:1339–1352
Kabata Z, Ho J (1981) The origin and dispersal of hake (genus
Merluccius: Pices: Teleostei) as indicated by its copepod
parasites. Oceanogr Mar Biol Annu Rev 19:381–404
Kartavtsev YP, Lee JS (2006) Analysis of nucleotide diversity
at the cytochrome b and cytochrome oxidase 1 genes at
the population, species and genus levels. Russ J Genet
42:341–362
Kartavtsev YP, Jung SO, Lee YM, Byeon HK, Lee JS (2007)
Complete mitochondrial genome of the bullhead torrent
catfish, Liobagrus obesus (Siluriformes, Amblycipididae):
genome description and phylogenetic considerations
inferred from the Cyt b and 16S rRNA genes. Gene
396:13–27
Knowlton N, Weight LA, Solorzano LA, Mills DK, Ber-
mingham E (1993) Divergence in proteins, mitochondrial
DNA, and reproductive compatibility across the Isthmus
of Panama. Science 260:1629–1632
Kocher TD, Thomas WK, Meyer A, Edwards SV, Paabo S,
Villablanca XF, Wilson AC (1989) Dynamics of mito-
chondrial DNA evolution in animals: amplification and
sequencing with conserved primers. Proc Natl Acad Sci
USA 86:6196–6200
Lankford TE, Billerbeck JM, Conover DO (2001) Evolution of
intrinsic growth and energy acquisition rates: II. Trade-
offs with vulnerability to predation in Menidia menidia.
Evolution 55:1873–1881
Legendre P, Legendre L (1998) Numerical ecology. Elsevier,
Amsterdam
Lloris D, Matallanas J, Oliver P (2003) Merluzas del mundo.
(Familia Merlucciidae). Catalogo comentado e ilustrado
de las merluzas conocidas. FAO Catalogo de especies
para fines de la pesca, Roma
Lozano-Cabo F (1965) Las merluzas atla
´nticas. Publ Te
´c J Est
Pesca 4:11–31
Lundy CJ, Moran P, Rico C, Milner RS, Hewitt GM (1999)
Macrogeographical population differentiation in oceanic
environments: a case study of European hake (Merluccius
merluccius), a commercially important fish. Mol Ecol
8:1889–1898
Martı
´nez E (1976) Variabilidad de los caracteres merı
´sticos en
Merluccius gayi (Guichenot, 1848) en la costa de Chile.
Rev Biol Mar 16:71–93
Mathews CP (1985) Meristic studies of the Gulf of California
species of Merluccius, with a description of a new species.
J Nat Hist 19:697–718
Mayr E (1969) Principles of systematic zoology. McGraw-Hill,
New York
McDowall RM (2008) Jordan’s and other ecogeographical
rules, and the vertebral number in fishes. J Biogeogr
35:501–508
McHugh J (1951) Meristic variations and populations of
northern anchovy (Engraulis mosdax). Scripps Inst Oce-
anogr Bull 6:123–160
Meyer A (1993) Evolution of mitochondrial DNA in fishes. In:
Hochachka PW, Mommsen TP (eds) Biochemistry and
molecular biology of fishes. Elsevier, Amsterdam, pp 1–38
Miller DJ (1956) Anchovy. Calif Coop Ocean Fish Invest Rept
1:20–26
Nei M, Kumar S (2000) Molecular evolution and phylogenet-
ics. Oxford University Press, New York
Nylander JAA (2004) MrModeltest v2. Program distributed by
the author. Evolutionary Biology Centre, Uppsala
University
Ovenden JR (1990) Mitochondrial DNA and marine stock
assessment: a review. Aust J Mar Freshwat Res 41:835–
853
Peregrino-Uriarte AB, Pacheco-Aguilar R, Yepiz-Plascencia G
(2007) Diferencias en los genes 16SrRNA y citocromo c
oxidasa subunidad I de las lisas Mugil cephalus yMugil
curema y los robalos Centropomus viridis yCentropomus
robalito. Cienc Mar 33:95–104
Posada D, Crandall KA (1998) Modeltest: testing the model of
DNA substitution. Bioinformatics 14(9):817–818
Quinn GP, Keough MJ (2002) Experimental design and data
analysis for biologists. Cambridge University Press,
Cambridge, UK
Quinteiro J, Vidal R, Rey-Me
´ndez M (2000) Phylogeny and
biogeographic history of hake (genus Merluccius) inferred
from DNA control-region sequences. Mar Biol 136:163–
174
Riddle BR, Hafner DJ, Alexander LF, Jaeger JR (2000) Cryptic
vicariance in the historical assembly of a Baja California
peninsular desert biota. PNAS 97:14438–14443
Rogers AR, Harpending H (1992) Population growth makes
waves in the distribution of pairwise genetic differences.
Mol Biol Evol 9:552–569
Rolda
´n MI, Garcı
´a-Marı
´n JL, Utter FM, Pla C (1998) Popu-
lation genetic structure of European hake, Merluccius
merluccius. Heredity 81:327–334
Rev Fish Biol Fisheries (2011) 21:259–282 281
123

Rolda
´n MI, Garcı
´a-Marı
´n JL, Utter FM, Pla C (1999) Genetic
relationships among Merluccius species. Heredity 83:79–
86
Rosel PE, Kocher TD (2002) DNA-based identification of
larval cod in stomach contents of predatory fishes. J Exp
Mar Biol Ecol 267:75–88
Rozas J, Sanche
´z-DelBarrio JC, Messenguer X, Rozas R
(2003) DnaSP, DNA polymorphism analyses by the coa-
lescent and other methods. Bioinformatics 19:2496–2497
Saitou N, Nei M (1987) The neighbor-joining method: a new
method for reconstructing phylogenetic trees. Mol Biol
Evol 4:406–425
Saunders MW, Leaman BM, McFarlane GA (1997) Influence
of ontogeny and fishing mortality on the interpretation of
sablefish, Anoplopoma fimbria, life history. NOAA Tech
Rept NMFS 130:81–92
Shields GF, Gust JR (1995) Lack of geographic structure in
mitochondrial DNA sequences of Bering Sea walleye
pollock, Theragra chalcogramma. Mol Mar Biol Bio-
technol 4:69–82
Short JA, Gburski CM, Kimura DK (2006) Using otolith
morphometrics to separate small Walleye Pollock Ther-
agra chalcogramma from Artic Cod Boregadus saida in
mixed samples. Alaska Fish Res Bull 12:147–152
Smith PE, Moser HG (2003) Long-term trends and variability
in the larvae of Pacific sardine and associated fish species
of the California Current region. Deep-Sea Res II
50:2519–2536
Smith PJ, Patchell GJ, Benson PG (1979) Glucosephosphate
isomerase and isocitrate dehydrogenase polymorphisms in
the hake, Merluccius australis. N Z J Mar Freshw Res
13:545–547
StatSoft (2004) STATISTICA, Version 7. Tulsa, OK
Swain DP (1992) The functional basis of natural selection for
vertebral traits of larvae in the stickleback Gasterosteus
aculeatus. Evolution 46:987–997
Swofford DL (2003) PAUP*. Phylogenetic analysis using
parsimony (*and other methods). Version 4. Sinauer
Associates, Sunderland
Tamura K, Nei M (1993) Estimation of the number of nucleotide
substitutions in the control region of mitochondrial DNA in
humans and chimpanzees. Mol Biol Evol 10:512–526
Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4:
Molecular Evolutionary Genetics Analysis (MEGA).
Software version 4.0. Mol Biol Evol 24:1596–1599
Teletchea F, Laudet V, Ha
¨nni C (2006) Phylogeny of the
Gadidae (sensu Svetovidov, 1948) based on their mor-
phology and two mitochondrial genes. Mol Phylogenet
Evol 38:189–199
Thompson JD, Higgins DG, Gibson TJ (1994) CLUSTAL W:
improving the sensitivity of progressive multiple sequence
alignment through sequence weighting, position-specific
gap penalties and weight matrix choice. Nucleic Acids
Res 22:4673–4680
Utter FM, Hodgins HO (1971) Biochemical polymorphisms in
the Pacific hake (Merluccius productus). Rapp P-V Reun
Cons Perm Int Explor Mer 161:87–89
Utter FM, Stormont CJ, Hodgins HO (1970) Esterase poly-
morphism in vitreous fluid of Pacific hake, Merluccius
productus. Anim Blood Grps Biochem Genet 1:69–82
Vrooman AM, Paloma PA (1977) Dwarf hake off the coast of
Baja California, Me
´xico. Calif Coop Ocean Fish Invest
Rep 19:67–72
Ward RD, Zemlak TS, Innes BH, Last PR, Hebert PDN (2005)
DNA barcoding Australia’s fish species. Phil Trans R Soc
B 360:1–11
Xia X, Xie Z (2001) DAMBE: data analysis in molecular
biology and evolution. J Hered 92:371–373
Yamahira K, Lankford TE Jr, Conover DO (2006) Intra- and
interspecific latitudinal variation in vertebral number of
Menidia spp. (Teleostei: Atherinopsidae). Copeia
2006:431–436
282 Rev Fish Biol Fisheries (2011) 21:259–282
123