228 
Fishery Bulletin 117(3) 
zone. The next largest group of exchange was the GOM, 
where 22% (77=20) of recaptures occurred. Cobia tagged 
in Brevard in January-March were largely recaptured in 
the GOM during April-May. Exchange also occurred with 
the south of Brevard and Florida Keys zones (77=14, 16% 
collectively), north of Brevard (n=10, 11%), and the zones 
north of the Florida border (77=13, 14% collectively). 
Florida Keys 
In a pattern similar to that of the Brevard zone, cobia 
were tagged ( 77 =181) and recaptured during every month 
of the year in this zone. The majority of tagging events 
(77=109,60%) and recaptures (77=113,56%) occurred during 
December-March. Only 18% (z? =32) of tagging events and 
21% (n=43) of recaptures occurred during June-October. 
Most fish tagged in the Florida Keys zone were recaptured 
in the same zone (77=104, 57%), often in close proximity 
to the tagging location (Fig. 3). The next greatest area of 
exchange was the GOM zone ( 77 = 66 , 37%), with most recap¬ 
tures occurring during April and May. Conversely, of the 
201 cobia recaptured in the Florida Keys zone, 45% (z?=90) 
were originally tagged in the GOM. Cobia tagged in the 
Florida Keys zone were recaptured in the GOM from Texas 
to the west coast of Florida, and recaptures from the Flor¬ 
ida Keys and GOM zones collectively account for 94% of all 
recaptures for this group. The remaining 6% ( 77 =10) were 
recaptured on the east coast of Florida (Brevard/south of 
Brevard zones). To date, no cobia tagged in the Florida 
Keys zone have been recaptured north of Cape Canaveral, 
and no cobia tagged north of Florida on the East Coast 
have been recaptured in the Florida Keys zone. 
Tagging summary 
Cobia tagged in the GOM were captured in the western 
North Atlantic Ocean (77=59, 6%) and cobia tagged in the 
western North Atlantic Ocean were recaptured in the GOM 
(77=26, 4%) in relatively small numbers. However, most of 
this exchange occurred between the GOM and the east 
coast of Florida. Movements of cobia tagged north of Florida 
to the GOM (t 7=4, 0.6%) and from the GOM to north of Flor¬ 
ida (?i=4, 0.4%) were very rare. Movements of fish tagged 
above the current stock delineation at the border of Flor¬ 
ida and Georgia to the north of Brevard and Brevard zones 
in northern and central Florida (77=21, 4%) and vice versa 
(77=14, 14%) were somewhat more common. Evaluating the 
frequency of recaptures by recapture location can provide 
information on the seasonal availability and susceptibil¬ 
ity of cobia to the fishery during the period of the study. In 
the northern area, from Georgia through the Mid-Atlantic 
region, cobia were mostly recaptured during May-August 
(77=434, 88%) with a peak in June (n=144, Table 6). In the 
GOM, cobia were present for a slightly longer period with 
recaptures mostly occurring between April and September 
(77=767, 84%), peaking in April (77=164). In contrast to the 
areas from Georgia to the north and the GOM, cobia recap¬ 
tures along the east coast of Florida and the Florida Keys 
occurred throughout the year, with no clear temporal peaks. 
Table 6 
Number of cobia (Rachycentron canadum) recaptured by 
month from 1988 through 2017 for 3 regions along the 
southeastern coast of the United States. GOM=Gulf of 
Mexico. 
East Florida and 
North of 
Month 
GOM 
Florida Keys 
Florida 
January 
11 
41 
2 
February 
12 
39 
1 
March 
28 
61 
1 
April 
165 
35 
18 
May 
141 
24 
103 
June 
139 
15 
144 
July 
127 
23 
112 
August 
111 
20 
75 
September 
85 
10 
26 
October 
58 
18 
7 
November 
26 
20 
3 
December 
11 
34 
3 
Genetics 
A total of 2796 samples meeting our selection criteria were 
successfully genotyped for inclusion in data analyses; col¬ 
lection years for samples included 2005 through 2017 
(Table 2). Only a single duplicate sample and 39 hatchery- 
produced fish occurred within the original data set. No 
large family groups (>3 siblings) were present within the 
data set, and only 12 fullsib pairs were identified (P=1.0); 
therefore, no confounding effects from family structure are 
anticipated in further analyses. 
Results from multiple rounds of hierarchical STRUC¬ 
TURE, initial pairwise •^'sT> and HWE analyses support 
the notion of a genetically distinct South Carolina inshore 
population and a homogenous GOM population ranging 
from Texas through the Ft. Pierce, Florida, area (FLE2) 
(Fig. 4). Additionally, the Virginia and inshore North Car¬ 
olina (NCI, NC2) samples represented a distinct genetic 
grouping, as did the combined offshore South Carolina 
and North Carolina samples (SCO, SCOl, SC02, NCOl, 
NC02) (Fig. 4). Samples from Cape Canaveral, Florida, 
through Savannah, Georgia, had genetic similarities 
with samples from collection locations both to the north 
(SCOs) and to the south (FLE2) and appeared to reflect 
a geographic transition zone in the STRUCTURE analy¬ 
ses (Fig. 4). As such, the iterative AMOVA were employed 
to evaluate potential breaks in gene flow within the area, 
including all potential locations from those of Atlantic 
Ocean offshore (SCOs, NCOs) and Savannah (GA) sam¬ 
ples through those of Jupiter Beach (FLEl) and Hobe 
Sound (FLE2) in Florida. Results indicate that the stron¬ 
gest significant break (jR st = 0.0073, P=0.001) among the 
groupings occurred with the separation between the Cape 
Canaveral (FLE3) and Jacksonville, Florida/Brunswick, 
Georgia (FLGA) locations, explaining 0.73% of the varia¬ 
tion in the data set. However, grouping scenarios between 
