Sme et al.: Identification of Eleginus gracilis by means of microsatellite markers 
61 
can waters, is little studied, but its position in the 
food web, potential population responses to warming 
and reduction of sea ice in the Arctic, and proposed 
offshore oil and gas development make learning about 
this species imperative. 
The distributions of several other gadid species— 
Arctic cod (B. saida), Pacific cod (Gadus macrocepha- 
lus), walleye pollock ( Gadus chalcogrammus), and 
Pacific tomcod ( Microgadus proximus) —overlap with 
that of E. gracilis, and furthermore navaga ( Eleginus 
nawaga) from the western Arctic Ocean is a congener 
of E. gracilis. Small gadids of several species are very 
similar morphologically and often present challenges 
for identification. The morphological bases of gadiform 
taxonomy, including the subfamily Gadinae to which 
all of the species in our study belong, have been de¬ 
scribed (e.g., Schultz and Welander, 1935; Svetovidov, 
1948; Cohen, 1989), as have the phylogenetic relation¬ 
ships among gadiform families (e.g., Roa-Varon and 
Ortf, 2009) and within Gadinae (Teletchea et al., 2006). 
However, questions remain about the relationships 
among E. gracilis, E. nawaga, and M. proximus (e.g., 
Carr et al., 1999; Roa-Varon and Ortf, 2009). Moreover, 
the modern geographic separation between E. eleginus 
and E. nawaga, if any exists, is unknown. 
Genetic analyses of a species can provide insight 
into several facets of its biology, including population 
structure, life history (e.g., Kamin et al., 2014), and 
recent demographic history (e.g., Harpending et al., 
1998). Information about population structure can be 
obtained from surveys in different geographic regions 
and the fish tested for genetic variation. Microsatellite 
data are beneficial, when compared with other classes 
of molecular markers, in that they are often highly 
polymorphic in fish species (DeWoody et al., 2000) and 
are relatively inexpensive to apply. Consequently, mi¬ 
crosatellite markers were isolated from and developed 
for E. gracilis. Here we 1) examine their variability 
in two E. gracilis collections from geographically sepa¬ 
rated areas; 2) determine their cross-reactivity with 
other northern Pacific and Arctic ocean gadids and the 
ability of suites of these loci to accurately distinguish 
among species; and 3) evaluate differences in the allele 
profiles among M. proximus, E. nawaga, and the two 
collections of E. gracilis. 
Materials and methods 
Samples and DNA isolation 
Collections of E. gracilis were collected from the 
Chuckchi Sea in 2011 and near Kodiak Island, Alaska, 
in 2013. Collections of B. saida from the Chukchi Sea 
were made in 2012 and collections of E. nawaga were 
collected from the Barents Sea in 2013. In 2015, G. 
chalcogrammus were collected in the southeast Ber¬ 
ing Sea. Two collections of M. proxi/nus were obtained, 
one from Puget Sound, Washington, between 1997 and 
1999, and one from Prince William Sound, Alaska, in 
2012. Two collections of G. microcephalus were collect¬ 
ed in 2013, one collected from Puget Sound and the 
other from Unimak Pass in the northern Gulf of Alaska 
(see details for all collections in Table 1). 
Tissue samples were preserved in a DNA preserva¬ 
tive solution (Seutin et al., 1991) or 95% ethanol and 
stored in the laboratory at -20°C. Total cellular DNA 
was isolated with Gentra Puregene 3 or Qiagen DNeasy 
kits (Qiagen, Hilden, Germany) by following the manu¬ 
facturer’s instructions. 
Discovery of microsatellites 
An Illumina paired-end shotgun library (Illumina, Inc., 
San Diego, CA) was prepared by shearing 1 pg of DNA 
from a single E. gracilis Chukchi Sea individual with 
a Covaris S220 focused-ultrasonicator (Covaris, Inc., 
Woburn, MA). The standard protocol for the TruSeq 
DNA library kit (Illumina, Inc.) and a multiplex iden¬ 
tifier adaptor index were used (see e.g., Stoutamore et 
al., 2012). A HiSeq system (Illumina, Inc.) was used 
to sequence 100-base pair [bp] paired-end readings. 
The program PAL_FINDER, vers. 0.02.03 (Castoe et 
al., 2012) was used to analyze 5 million of the result¬ 
ing sequences to identify readings that had di-, tri-, 
tetra-, penta-, and hexanucleotide repeat motifs. The 
data are archived in the Sequence Read Archive of the 
National Center for Biotechnology Information under 
accession number SAMN06333955. Once positive reads 
were identified, oligonucleotide primers were designed 
with the program PrimerS, vers. 2.0.0 (Koressaar and 
Remm, 2007; Untergasser et al., 2012). To avoid issues 
with copy number of primer sequences in the genome, 
loci for which the primer sequences occurred only once 
or twice in the 5 million reads were selected. Forty- 
eight presumed loci from E. gracilis that met this cri¬ 
terion were chosen for primer design. 
The 48 primer pairs were tested with DNA from 8 
E. gracilis individuals. The polymerase chain reactions 
(PCRs) were conducted over two 10°C spans of anneal¬ 
ing temperatures (65-55°C or 58-48°C) with touch¬ 
down thermal cycling profiles (Don et al., 1991). The 
results (not presented) were analyzed with GeneMap- 
per, vers. 3.7 (Thermo Fisher Scientific, Waltham, MA). 
Eighteen primer pairs were then selected for evalua¬ 
tion with larger sample sizes. 
Analysis of microsatellites 
Target sequences of the 18 primer pairs amplified with 
a touchdown PCR strategy reduced nontarget bands 
in the product spectrum (Don et al., 1991). All reac¬ 
tions contained ~ 1 unit Taq polymerase, lx PCR buf¬ 
fer (50 mM KC1 2 , 10 mM Tris-HCl pH 9.0, 0.1% Triton 
X-100; Promega Corp., Madison, WI), 0.5 pM deoxyri- 
bonucleotide triphosphates, and 0.025 to 0.1 pg DNA 
3 Mention of trade names or commercial companies is for iden¬ 
tification purposes only and does not imply endorsement by 
the National Marine Fisheries Service 
