Lynch et al.: A genetic investigation of the population structure of Brevoortici tyrannus 
89 
coast regions in 2007 to corroborate field identifications. 
Some of the identifications were retroactively verified 
with sequences from a portion of the mtDNA control 
region, namely haplotypes were compared to those of 
Anderson (2007). 
Total genomic DNA was extracted from each tissue 
sample by using a Qiagen DNeasy® Tissue Kit (Qia- 
gen, Valencia, CA) and following the manufacturer’s 
protocol. A 459 base-pair (bp) fragment of the mito- 
chondrial cytochrome c oxidase subunit I (COI) gene 
region (F: 5'CTTTCGGCTACATGGGAATG3' and R: 
5AGCCCTAGGAAGTGTTGTGG5', GenBank acces- 
sion number DQ867533), a 535-bp fragment of the mi- 
tochondrial control region (Pro-F: 5' CTA CCY CYA 
ACT CCC AAA GC 3', [Gray et al., 2008] and Phe-R: 
5' GTA AAG TCA CGA CCA AAC C 3', [Brendtro et 
al., 2008], and eight microsatellite loci Asa2, Asa4, 
Asal6, [Brown et al., 2000]; Aal6, [Faria et al., 2004]; 
AsaB020, AsaC334, AsaD055, [Julian and Bartron, 
2007]; SarBH04, [Pereyra et al., 2004]), were amplified 
in either 5-pL (microsatellites) or 10-pL (COD reaction 
volumes using the polymerase chain reaction (PCR) 
with the conditions outlined in Lynch (2008). 
Mitochondrial PCR products were purified for se- 
quencing by using column filtration with a QIAquick® 
PCR purification kit (Qiagen) following the manufactur- 
er’s protocol, and the concentration was measured with 
a BioMate™ 3 series UV Sspectrophotometer (Thermo 
Spectronic, Madison, WD. PCR products were prepared 
for sequencing with the ABI PRISM® BigDye™ Termi- 
nator, vers 3.1 cycle sequencing kit (Applied Biosystems, 
Foster City, CA) at a 1:8 dilution and sequenced on an 
80-cm capillary ABI PRISM® 3130xZ genetic analyzer 
(Applied Biosystems). Samples were sequenced in the 
forward and reverse direction. 
The chromatographic curves for each 80-cm capil- 
lary sequence were analyzed using Sequencing Analy- 
sis Software, vers. 5.2 (Applied Biosystems). All mito- 
chondrial sequences were edited with Sequencher 4.7.2 
(Gene Codes Corp., Ann Arbor, MI), variable positions 
were confirmed visually, and sequences were aligned by 
using the ClustalW algorithm (Thompson et al., 1994) 
for multiple alignments in MacVector 9.0.1 (MacVector 
Inc., Cary, NC). 
Microsatellite loci were amplified by PCR by using 
locus-specific fluorescent labels with the conditions out- 
lined in Lynch (2008). Following amplification, 1 pL of 
PCR product for each locus was combined with PCR 
products from three other unique locus and fluorescent 
label combinations (4 pL total), 6 pL HiDi formamide 
(Applied Biosystems), and 0.3 pL 500 Liz Gene Scan 
Size standard (Applied Biosystems). The reaction mix- 
ture was denatured at 95°C for 10 minutes before being 
separated on a 36-cm capillary ABI PRISM® 3130xZ 
Genetic Analyzer (Applied Biosystems) according to 
the manufacturer’s protocol. The chromatic peaks for 
each microsatellite locus were scored by GeneMarker 
AFLP/Genotyping Software, vers. 1.60 (SoftGenetics, 
State College, PA). Once scored, MicroChecker 2.2.3 
(Van Oosterhout et al., 2004) was used to check for 
the presence of null alleles and evidence of scoring er- 
rors. To ensure consistency, 20% of the samples were 
re-analyzed from the point of DNA extraction through 
allele scoring. 
Genetic analyses 
Once aligned, the mitochondrial sequences were char- 
acterized in Arlequin 3.11 (Excoffier et al., 2005) to 
determine the number of haplotypes (N h ), number of 
polymorphic sites (S), and variable base-pair (bp) loca- 
tions within a sequence set. Diversity indices, including 
haplotype diversity (h), nucleotide sequence diversity (jr), 
and mean number of pairwise differences ( k ) within each 
collection were also estimated in Arlequin 3.11 (Excoffier 
et al., 2005). To visualize genetic relationships among 
mitochondrial sequences, median-joining networks were 
drawn in Network 4. 2. 0.1 (Bandelt et al., 1999). 
For the microsatellite data, Genepop 3.4 (Raymond 
and Rousset, 1995) was used to determine observed 
heterozygosity (H Q ) and expected heterozygosity (H E ) 
and to perform exact tests for deviations of genotypic 
distributions from the expectations of Hardy-Weinberg 
equilibrium for each locus at each collection location 
(10,000 iterations; Guo and Thompson, 1992). Signifi- 
cance levels were adjusted for multiple testing by using 
a Bonferroni correction (Rice, 1989). Arlequin 3.11 (Ex- 
coffier et al., 2005) was used to determine the number 
of alleles (a), and Microsatellite Analyzer (MSA) (Di- 
eringer and Schlotterer, 2003) was used to determine 
the allele size range (as). Allelic richness (R s ) was esti- 
mated in FSTAT 2. 9. 3. 2 (Goudet, 1995). 
Using both mitochondrial COI sequence data (<f> S7 4 and 
nuclear microsatellite data ( F ST /R ST ), we performed a 
hierarchical analysis of molecular variance (AMOVA) to 
test for partitioning of variation among defined groups. 
The groups tested were the following: temporal collec- 
tions within an age class at a location (e.g., 2007 YOY in 
Chesapeake Bay sampled early [May] and late [August] 
in the season), between collections of an age class taken 
at the same location in different years (e.g., the 2006 
year class sampled in Chesapeake Bay as YOY in 2006 
and yearling in 2007), between age classes within a 
region (e.g., YOY and yearling menhaden in Chesapeake 
Bay in 2007), among Atlantic coast regions both includ- 
ing and excluding the Gulf of Mexico (e.g., New England, 
mid-Atlantic, Chesapeake Bay, U.S. South Atlantic, and 
Gulf of Mexico), among COI clades (e.g., “Atlantic only,” 
“ubiquitous,” and “anomalous” samples), and between 
Atlantic and Gulf menhaden. AMOVA calculations based 
on microsatellite data were analyzed by using both ^st 
(Weir and Cockerham, 1984) and R ST (Slatkin, 1995) 
distance methods. Estimates of population pairwise <P ST 
and F st /R st were calculated in Arlequin 3.11 (Excoffier 
et al., 2005) and adjusted for multiple testing with a 
Bonferroni correction (Rice, 1989). 
To assess the statistical power for detecting popula- 
tion differentiation with the applied set of microsatellite 
markers and sample sizes, a simulation was imple- 
mented with POWSIM (Ryman and Palm, 2006), which 
