Southeastern Naturalist
C.A. Kalinowsky, M.C. Curran, and J.W. Smith
2016 Vol. 15, No. 4
714
2016 SOUTHEASTERN NATURALIST 15(4):714–728
Age and Growth of Rachycentron canadum (L.) (Cobia) from
the Nearshore Waters of South Carolina
Christopher Aaron Kalinowsky1,*, Mary Carla Curran2, and Joseph W. Smith3
Abstract - The purpose of this study was to define growth parameters, age-at-length, and
the sex ratio for Rachycentron canadum (Cobia) in Port Royal Sound and the nearshore
waters of South Carolina. We sampled Cobia from recreational-fishing efforts, and used
otoliths to estimate age. Female Cobia (n = 245) fork length (FL) ranged from 798 mm to
1425 mm (mean = 1059 mm) and male (n = 221) FL ranged from 670 mm to 1183 mm (mean
= 936 mm). The ratio of females to males was 1.1:1.0. Cobia ranged in age from 2 to 11
years; most (60.8%) were age 3. Estimates of von Bertalanffy growth parameters for Cobia
were L∞ = 1212, K = 0.53, and t0 = -0.13 for females and L∞ = 1101, K = 0.51, and t0 = -0.13
for males. Life-history characteristics of Cobia as defined by this study provide managers
with critical age-at-length and growth information necessary for the effective management
of the species.
Introduction
Rachycentron canadum (L.) (Cobia) are large, coastal, pelagic fish of the
monotypic family Rachycentridae that are distributed worldwide in tropical and
subtropical seas, except the eastern Pacific (Herre 1953, Robins and Ray 1986,
Shaffer and Nakamura 1989). Along the east coast of the US, Cobia occur in nearshore
waters from the Mid-Atlantic Bight to the Gulf of Mexico (Williams 2001).
This Cobia population is divided into 2 regional stocks (Gulf and Atlantic), each
of which is managed by its representative fisheries management council under
the Coastal Pelagics Management Plan (SEDAR 2013a, b; Shaffer and Nakamura
1989; Williams 2001). Migration of individuals between the Gulf of Mexico
and the Atlantic Ocean has been documented through tagging studies, and early
genetic studies (using mtDNA analysis) indicated that the 2 stocks are genetically
similar (Franks et al. 1991, Howse et al. 1992, Hrincevich 1993). However,
more-recent studies have found that the 2 stocks have disparate allele-frequency
distributions, indicating some degree of isolation between the stocks (SEDAR
2013a, b). The Gulf of Mexico stock extends around the tip of Florida as far north
as Brevard County, where some degree of overlap occurs with the Atlantic stock.
Currently, there is not enough resolution in the genetic or tagging studies to identify
exactly where the 2 stocks split. Genetic analysis indicates that a mixing zone
occurs somewhere to the north of the Brevard County line (SEDAR 2013a, b). For
1Georgia Department of Natural Resources-Coastal Resources Division, 185 Richard Davis
Drive Suite 104, Richmond Hill, GA 31324. 2Department of Marine and Environmental
Sciences, Box 20467, Savannah State University, Savannah, GA 31404. 3NOAA/NMFS,
101 Pivers Island Road, Beaufort, NC 28516. *Corresponding author - Chris_Kalinowsky@
dnr.state.ga.us.
Manuscript Editor: Benjamin Keck
Southeastern Naturalist
715
C.A. Kalinowsky, M.C. Curran, and J.W. Smith
2016 Vol. 15, No. 4
management purposes, it was decided that the stocks would be separated at the
FL/GA line (SEDAR 2013a, b).
Cobia are a highly prized gamefish, and a significant recreational fishery exists
for them in the US (Hammond 2001). It was estimated that over 2.4 million
Cobia were caught recreationally on the Gulf and Atlantic coasts from 1989 to
2000 (Hammond 2001). A majority of these landings (72%) occurred along the
Gulf Coast; more specifically, 55% of the landings occurred along the west coast
of Florida (Hammond 2001). Commercial harvest of Cobia is limited and occurs
mostly as by-catch from other fisheries (Shaf fer and Nakamura 1989).
Stock status of Cobia in the Gulf of Mexico and the Atlantic remains uncertain.
This uncertainty is largely the result of limited data on migration patterns, overwintering
locations, and population estimates. A 2001 stock assessment from the Gulf
of Mexico had a high degree of uncertainty due to the lack of landings data and
gaps in general life-history information (Williams 2001). The results from that assessment
were highly variable and suggested that the population status could range
from overfished to well above maximum sustainable yield (Williams 2001). Based
on the results of the 2001 assessment, Williams (2001) stated that the population of
Gulf Cobia had increased since the 1980s and comprehensive coast-wide sampling
and ageing of Cobia would be necessary to improve future population estimates.
The most recent stock assessment, conducted in 2012, also contained a high degree
of uncertainty around estimates for both the Gulf and Atlantic stocks. The current
assessment continues to lack consistent data regarding many aspects of Cobia
life history and the fisheries that impact the species. Gaps in the data include the
identification of stock boundaries along the Western Central Atlantic, release mortality
for both commercial and recreational fisheries, spawning location, and size at
maturity. Despite these many uncertainties, results indicated that neither stock was
overfished and that overfishing was not occurring (SEDAR 2013a, b ).
Despite the popularity of Cobia, little is known about the status of the fishery
or the life history of the species along the South Carolina coast (Hammond 2001).
The purpose of the present study was to better define life-histo ry characteristics of
Cobia so that managers will have the age and growth information required to accurately
manage the species. Port Royal Sound in Beaufort County, SC, is among
several areas along the Western Central Atlantic Coast where Cobia move inshore
in spring to spawn (Lefebvre 2009). Anglers in Beaufort County land ~80–85% of
Cobia caught in South Carolina waters (Hammond 2001); thus, our objectives were
to determine growth parameters, age-at-length, and the sex ratio for Cobia in Port
Royal Sound and the adjacent waters of Beaufort County , SC.
Field-site Description
We sampled Cobia from recreational catches originating in Port Royal Sound
and the nearshore waters of Beaufort County, SC (Fig.1). Port Royal Sound is a
large, deep, highly saline system that has no significant source of freshwater input.
Port Royal Sound is bordered by St. Helena Island to the north and Hilton Head
Island to the southwest. Moving upriver, the sound splits into 3 distinct channels—
Southeastern Naturalist
C.A. Kalinowsky, M.C. Curran, and J.W. Smith
2016 Vol. 15, No. 4
716
the Beaufort River to the north, the Broad River in the middle, and the Chechessee
River to the south. The Broad River is the deepest and lar gest of the three.
Methods
Sample collection
We sampled Cobia opportunistically from the Port Royal Sound and adjacent
waters of Beaufort County (Fig. 1). Specimens (n = 485) were collected from a
targeted recreational fishery, including recreational anglers, charter boats, and local
sportfishing tournaments during May, June, and July 2005–2007. We neither
employed additional methods to collect undersized Cobia nor interviewed anglers
about discarded, undersized fish. We collected Cobia samples from volunteer recreational
anglers; thus, there was no way to measure effort or control sampling
Figure 1. Map of Port Royal Sound and the adjacent waters of Be aufort County, SC.
Southeastern Naturalist
717
C.A. Kalinowsky, M.C. Curran, and J.W. Smith
2016 Vol. 15, No. 4
methods of the fishermen. We used a measuring board to determine the fork length
(FL) of each specimen to the nearest millimeter. When available, we recorded
whole weight to the nearest ounce using a calibrated digital scale and then converted
the values to grams. We determined sex for all specimens by gross examination
of reproductive tissue.
Otolith sectioning
We removed sagittal otoliths in the field by opening the cranium above and
slightly posterior to the preoperculum. Once removed, otoliths were rinsed in fresh
water and stored dry in labeled coin envelopes or plastic vials. Prior to sectioning,
we marked otolith cores (focus of the otolith) with a fine-point marker and
then embedded whole otoliths in a two-part epoxy in a small (14-mm) bullet mold
(Pelco #10504, Pelco Tool and Mold, Glendale Heights, IL). We placed molds with
embedded otoliths concave side down on microscope slides, secured them with
cyanoacrylate glue, and cross-sectioned them through the focus using a Buehler
low-speed saw equipped with a diamond-edged wafering blade (Buehler, Lake
Bluff, IL). We mounted the otolith sections onto a microscope slide with an acrylic
resin (CytoSeal 60, Thermo Fisher Scientific, Waltham, MA) and viewed them with
a dissecting microscope at 16x using transmitted light to determine the quality of
the section. If a second section was deemed necessary, we remounted the remaining
half of the bullet mold, cut a section adjacent to the first cut, and mounted it
on the same slide as the original. We viewed finished slides under a Leica S8APO
stereomicroscope (10–80x) using fluorescent light to discern and count annuli.
Ageing methods
Three experienced readers independently examined sectioned otoliths, and assigned
age estimates by counting opaque zones along the mid-portion of the ventral
lobe per Burns et al. (1998), Franks et al. (1999), Smith (1995), and Thompson et
al. (1991). Ages were assigned without reference to fork length, weight, or date
caught. All otoliths had a wide band of translucent material extending toward the
terminal edge from the prior year’s annuli. Most otoliths also had a distinct opaque
zone at the terminal edge of the section, making age determination straightforward
and resulting in strong agreement among readers (97.3%). However, there were a
few otoliths (n = 13) that had a faint, but still visible, opaque zone that was not as
distinct as the prior year’s annuli. Those otoliths accounted for the majority of disagreements
in age determination among readers. In these instances, we determined
that any opaque zone present along the terminal edge indicated that the annulus was
forming at the time of capture and it was counted as another year. For example, a
fish was assigned an age of 3 if it had 2 opaque zones with a large translucent zone
and faint opaqueness at the terminal edge. Those few problematic otoliths were
reviewed and ages reassigned. In all cases of reader disagreement, at least 2 of the
3 readers independently assigned the same age during their initial determination.
When reader-assigned ages differed, readers re-examined otoliths independently
and re-assigned ages. We excluded 2 otoliths from the study because readers could
not agree on an age. Franks et al. (1999) found that age estimates from both the left
Southeastern Naturalist
C.A. Kalinowsky, M.C. Curran, and J.W. Smith
2016 Vol. 15, No. 4
718
and right otoliths agreed when compared; however, to enhance consistency in the
present study, we always used the left otolith unless it was damaged.
Analysis of length
We sorted fork lengths of specimens into 50-mm bins (range = 600–1450 mm
FL) and constructed length-frequency distributions by sex. We employed the Kolmogorov-
Smirnov 2-sample test to compare the length-frequency distributions
of males and females by using the NPAR1WAY procedure in SAS (SAS Institute
Inc. 1999).
For the parameter estimates observed, age and length data were fitted to the von
Bertalanffy growth model using the least-squares method. We chose this method
because it is widely used in fisheries studies (Haddon 2001). The Von Bertalanffy
growth model (lt = L∞(1 - e-K[t - t0]) is a 3-parameter equation, where lt is the length
at age t, L∞ is the asymptotic maximum length, K is the growth rate coefficient that
determines how quickly maximum size is reached, and t0 is the hypothetical age at
which the species has zero length (Haddon 2001).
To compensate for the lack of age-1 Cobia, we applied the Diaz-adjusted von
Bertalanffy method to generate alternative parameter-estimates. This method assumes
a censored-normal distribution for variability of observed lengths about
predicted lengths in the nonlinear regression of length-at-age. Age data were
truncated for younger fish because of harvest regulations prohibiting the take of
fish smaller than 838 mm FL (i.e., only the larger fish were sampled from younger
age classes because of the minimum-size-limit (838-mm) FL regulation. We made
a third attempt to further refine parameter estimates by using the Diaz-adjusted
model with a fixed t0 value of -0.13 for males. This procedure was based on the
assumption that t0 should be similar for both females and males, and that initial estimates
of t0 for males were skewed because of: (1) the limited age-2 data used for
the initial model and (2) the data restrictions inherent to the Diaz-adjusted model.
Analysis of growth
Cobia have been documented to have sex-specific growth patterns in North
Carolina and the Gulf of Mexico; therefore, we fitted growth models to sex-specific
data. We generated initial parameter-estimates by fitting observed length and age
data to the standard von Bertalanffy growth equation (Ricker 1975) and we obtained
secondary estimates by fitting age–length data to a Diaz-adjusted von Bertalanffy
growth equation (Diaz et al. 2004). We employed this method to compensate for
truncated age-1 data resulting from the regulatory minimum-size limit for Cobia.
We generated a third model using the standard von Bertalanffy techniques that included
non-truncated age-1 data from fish (n = 9) captured in a previous study in
North Carolina (Smith 1995). We intended to use this third model to compare the
results of the Diaz method to results using actual age-1 data to determine if a better
growth curve and a more accurate t0 could be produced. We fitted the age data
to a standard von Bertalanffy model for both sexes and used likelihood ratio tests
(Kimura 1980) to compare the results of this third model to determine if there were
significant diferrences in the growth curves of males and female s.
Southeastern Naturalist
719
C.A. Kalinowsky, M.C. Curran, and J.W. Smith
2016 Vol. 15, No. 4
Results
Over all collection years, Cobia (n = 485) ranged in size from 670 mm to 1425
mm FL (mean = 999 mm; Table 1). Female Cobia (n = 245) attained a greater maximum
size than males and ranged from 798 mm to 1425 mm FL (mean = 1059 mm);
males (n = 221) ranged from 670 mm to 1183 mm FL (mean = 936 mm; Table 1).
Unsexed specimens (n = 19) were recorded and used for combined age analysis and
combined length analysis.
Analysis of length
Length-frequency distributions for females and males were significantly different
(Kolmogorov-Smirnov 2-sample test: d = 0.47, P < 0.05), with female Cobia
larger than males (Wilcoxon-Mann-Whitney, P < 0.05; Fig. 2). The sex ratio of
females to males was 1.1:1. Modal length range for both sexes combined was 950–
1000 mm FL, accounting for 17.6% of all samples. The modal size was 1000–1050
mm FL for females (20.2% of all females and 10.3% for all fish) and 900–950 mm
FL for males (32.0% of all males and 14.7% of all specimens).
Table 1. Size range (mm FL) and mean length (mm FL ± 1 SE) by sex and by year (2005–2007) for
Rachycentron canadum (L.) (Cobia; n = 485) from Port Royal Sound and the adjacent waters of Beaufort
County, SC. Unsexed specimens (n = 19) were used to calculate all Cobia means (last column).
Female Male All Cobia
Year n Range Mean n Range Mean n Range Mean
2005 79 819–1410 1119 ± 15 39 855–1158 1002 ± 14 119 819–1410 1081 ± 12
2006 23 840–1236 1095 ± 23 34 811–1176 968 ± 17 57 811–1236 1019 ± 16
2007 143 798–1425 1019 ± 9 148 670–1183 910 ± 5 309 670–1425 964 ± 6
Combined 245 798–1425 1059 ± 8 221 670–1183 936 ± 6 485 670–1425 999 ± 6
Figure 2. Length-frequency diagrams for Rachycentron canadum (L.) (Cobia) collected
between 2005 and 2007 from Port Royal Sound and the adjacent waters of Beaufort County,
SC. Females (n = 245) and males (n = 221).
Southeastern Naturalist
C.A. Kalinowsky, M.C. Curran, and J.W. Smith
2016 Vol. 15, No. 4
720
Analysis of age
We examined sectioned annual zonal deposition of sagittal otoliths from 474
specimens. Estimates of Cobia ages ranged from 2 y to 11 y based on counts of
opaque zones on sectioned otoliths (Fig. 3). Modal age for females (54.2%) and
males (66.6%) was age 3 (Fig. 3). Ranges of FL by age showed considerable overlap
among most age groups for both sexes. For example, age-2 females ranged from
798 mm to 905 mm FL, while age-3 females ranged from 839 mm to 1160 mm FL.
Analysis of growth
The 2 sexes had significantly different growth curves (χ2 = 66.13, df = 3, P less than
0.001). Therefore, we separated length-at-age data by sex and used them to create sex
specific age–length tables (Table 2). Initial parameter estimates obtained from the
von Bertalanffy growth function were L∞ = 1425 mm FL (the asymptotic maximum
length), K = 0.21 (growth rate coefficient), t0 = -2.48 (modeling artifact used to represent
age when average length was zero) for females and L∞ = 1213, K = 0.21, t0 = -3.45
for males (Fig. 4A). Secondary-parameter estimates using the Diaz-adjusted von
Bertalanffy function were L∞ = 1212 mm FL, K = 0.53, t0 = -0.13 for females and L∞ =
1173 mm FL, K = 0.28, t0 = -2.14 for males (Fig. 4B). We ran the Diaz-adjusted model
a second time using a fixed t0 value of -0.13 for males, and the parameters for males
were L∞ = 1101 mm FL, K = 0.51 (Fig. 4C). For comparative purposes, we made a
fourth and final attempt to refine parameter estimates using age-1 data (n = 9) obtained
from an ageing study in North Carolina (Smith 1995). These data produced the
following parameter estimates: L∞ = 1358 mm FL, K = 0.33, t0 = -0.68 for females;
and L∞ = 1155 mm FL, K = 0.42, t0 = -0.51 for males (Fig. 4D). This method provided
the most realistic estimates for all parameters (Fig. 4D).
In general, the predicted age-at-length model developed from the von Bertalanffy
formula provided a good fit to the data. The fit was best for those age groups
Figure 3. Age-frequency diagrams for Rachycentron canadum (L.) (Cobia) collected between
2005 and 2007 from Port Royal Sound and the adjacent waters of Beaufort County,
SC. Females (n = 245) and males (n = 221).
Southeastern Naturalist
721
C.A. Kalinowsky, M.C. Curran, and J.W. Smith
2016 Vol. 15, No. 4
Table 2. Age–length table for female (n = 236) and male (n = 214) Rachycentron canadum (L.) (Cobia) from Port Royal Sound and the adjacent waters of
Beaufort County, SC, as compared to Smith (1995). Lengths were predicted using a standard von Bertalanffy growth model.
Present Study (SC) NC Study (Smith 1995)
Age Range observed Mean observed Von Bertalanffy Age Range observed Mean observed Von Bertalanffy
Sex (y) n FL (mm) FL (mm) ± 1 SE FL (mm) (y) n FL (mm) FL (mm) ± 1 SE FL (mm)
♀ 1 0 0 0 0 1 3 490–630 550 ± 40 610
2 8 798–905 855 ± 14 871 2 18 570–1060 810 ± 30 770
3 127 839–1160 979 ± 5 977 3 50 790–990 890 ± 10 890
4 22 935–1180 1068 ± 9 1061 4 23 880–1320 1020 ± 20 990
5 21 1061–1204 1133 ± 9 1131 5 13 980–1130 1060 ± 10 1070
6 20 1070–1294 1180 ± 11 1187 6 20 990–1260 1110 ± 20 1130
7 24 998–1381 1226 ± 16 1232 7 11 1100–1260 1170 ± 20 1170
8 2 1194–1373 1284 ± 90 1269 8 8 1140–1280 1230 ± 20 1210
9 8 1161–1410 1308 ± 35 1299 9 7 1140–1340 1250 ± 20 1240
10 2 1271–1356 1314 ± 43 1323 10 3 1170–1330 1270 ± 50 1260
11 2 1314–1425 1370 ± 56 1342 11 1 1210 1210 1280
12 0 - - - 12 3 1250–1300 1270 ± 20 1300
13 0 - - - 13 2 1340–1420 1380 ± 40 1310
14 0 - - - 14 0 - - -
♂ 1 0 0 0 0 1 6 390–640 500 ± 40 560
2 5 811–844 829 ± 6 825 2 22 630–930 740 ± 20 710
3 143 670–1010 897 ± 4 898 3 41 680–1020 820 ± 10 820
4 16 844–1095 956 ± 17 957 4 32 820–970 880 ± 10 890
5 18 950–1095 1020 ± 10 1005 5 20 780–990 920 ± 10 940
6 12 938–1158 1058 ± 17 1045 6 7 900–1030 950 ± 20 970
7 13 934–1104 1042 ± 13 1076 7 6 940–1080 1000 ± 20 1000
8 3 1078–1176 1116 ± 30 1102 8 8 890–1070 990 ± 20 1010
9 2 1150–1151 1151 ± 1 1123 9 6 990–1360 1070 ± 60 1030
10 1 1163 1163 1140 10 5 1010–1090 1050 ± 10 1030
11 1 1183 1183 1153 11 3 1020–1090 1050 ± 20 1040
12 0 - - - 12 0 - - -
13 0 - - - 13 1 - 1130 1040
14 0 - - - 14 1 - 1060 1050
Southeastern Naturalist
C.A. Kalinowsky, M.C. Curran, and J.W. Smith
2016 Vol. 15, No. 4
722
Figure 4. Von Bertalanffy
growth curves
for female and male
Rachycentron canadum
(L.) (Cobia) from
Port Royal Sound and
the adjacent waters
of Beaufort County,
SC (2005–2007) (x =
female, o = male). (A)
Using standard von
Bertalanffy methods.
(B) Using Diaz-adjusted
von Bertalanffy
methods. (C) Using
standard von Bertalanffy
methods with
fixed male t0 values.
(D) Using standard
von Bertalanffy methods
including actual
age-1 fish from NC
(Smith 1995).
with the most samples. For example, observed mean FL for age-3 females (n = 127)
was 979 ± 5 mm, which encompasses the predicted FL of 977 mm based on the von
Bertalanffy model (Table 2). Mean FL for age 3 males (n = 143) was 897 ± 4 mm,
which also encompassed the predicted FL of 898 mm (Table 2). Even in those age
groups with the fewest samples, there was only a 28-mm difference for females at
age 11 (n = 2) and 30 mm difference for males at age 11 (n = 1), or 2%, between
observed and predicted FLs (Table 2).
Southeastern Naturalist
723
C.A. Kalinowsky, M.C. Curran, and J.W. Smith
2016 Vol. 15, No. 4
Discussion
This study provides critical age and growth information for Cobia from South
Carolina, and the growth parameters generated from this study will aid in the
management of Cobia in the western central Atlantic. Prior to this study, managers
had little information on how Cobia from this region compared to those in others.
Our data show that Cobia from Port Royal Sound and adjacent waters of Beaufort
County, SC, have similar size ranges and growth rates as those from other nearby
regions. Smith (1995) found that female Cobia from North Carolina waters ranged
from 490 mm to 1420 mm FL (mean = 1090 mm), and males ranged from 390 mm
to 1360 mm FL (mean = 940 mm) (Table 3). Although similar in mean length, both
sexes of Cobia from North Carolina (Smith 1995) had a wider FL range than the
Cobia we sampled for our study (Table 3). This difference was most likely due to a
greater availability of samples accessible to Smith (1995), who obtained specimens
from a wide variety of non-selective gear including trawls, gill nets, pound nets,
stop nets, long hauls, and purse seines. These gear types typically recruit a larger
range of specimen sizes and ages. Mean lengths from the present study are also
similar to those reported by Franks et al. (1999), who found that female Cobia from
the Gulf of Mexico ranged from 335 mm to 1651 mm FL (mean = 1050 mm), and
Table 3. Regional parameter estimates for the von Bertalanffy growth model for Rachycentron canadum
(L.) (Cobia) from the present study and a table taken from Franks et al. (1999). Differences in parameter
estimates may be the result of different aging techniques (scales vs. otoliths) or different size
ranges due to sampling methodology, sample size, and/or lack of age-1 data. F = female and M = male.
Size range
Region Sex FL (mm) n L∞ k t0 Structure Authors
Virginia F 582–1377 156 1640 0.23 -0.08 Scales Richards 1967
M 544–1194 101 1210 0.28 -0.06
North Carolina F 490–1420 92 1350 0.24 -1.53 Otoliths Smith 1995
M 390–1360 116 1050 0.37 -1.08
Western Louisiana F 358–1445 - 1294 0.56 0.11 Otoliths Thompson et al.
M 528–1432 - 1132 0.49 -0.49 1991
Northeastern Gulf of F 335–1651 395 1555 0.27 -1.25 Otoliths Franks et al. 1999
Mexico M 345–1450 170 1170 0.43 -1.15
South Carolina
Initial von Bertalanffy F 798–1425 236 1425 0.21 -2.48 Otoliths Present study
estimates M 670–1183 214 1213 0.21 -3.45
Initial Diaz-adjusted F 840–1425 233 1212 0.53 -0.13 Otoliths Present study
estimates M 844–1183 209 1173 0.28 -2.14
Secondary Diaz- F 840–1425 233 1212 0.53 -0.13 Otoliths Present study
adjusted M 844–1183 209 1101 0.51 -0.13
Estimates using NC F 490–1425 239 1358 0.33 -0.68 Otoliths Present study
age-1 M 500–1183 220 1155 0.42 -0.51 (Including age-1
fish from Smith
[1995])
Southeastern Naturalist
C.A. Kalinowsky, M.C. Curran, and J.W. Smith
2016 Vol. 15, No. 4
724
males ranged from 345 mm to 1450 mm FL (mean = 952 mm). Franks et al. (1999)
also had access to more juvenile fish, which had been confiscated as undersized
specimens but made available for their use. The similarity in growth rates and
lengths between these different regions is interesting given the vast differences in
habitat type between the areas. This result may be due primarily to the overwintering
preferences of the species and the fact that fish migrate south or east in search
of warmer waters during the winter.
The pattern of annulus formation in Cobia from South Carolina that we observed
may result from slowed growth during migration to the coastal spawning areas.
This type of annulus formation has been documented in other species (Chiang et
al. 2004, Sun et al. 2002, Tserpes and Tsimenides 1995). The presence of an opaque
zone along the terminus of most otoliths examined in the present study supports
this theory. The opaque zones form before the fish appear in coastal waters of South
Carolina, thus suggesting that they are the result of slowed growth during migration
to the region. Our findings are similar to those observed by Franks et al. (1999),
who surmised that Cobia from the northeastern Gulf of Mexico formed their opaque
zones during migration into that region. These authors based this conclusion on the
presence of a terminal opaque zone on otoliths from Cobia sampled in April and in
sexually immature fish. Because these fish were just entering the area or were not
yet sexually mature, Franks et al. (1999) presumed that the terminal bands were
formed during the migration to the region rather than as a result of the spawning
event. Furthermore, other studies by Burns et al. (1998), Smith (1995), and Thompson
et al. (1991) utilized otoliths to age Cobia, and all documented a singular annulus
formation in the spring–summer months. Therefore, in the present study, age
for Cobia is presumed to be equal to the number of opaque zones present in the
sectioned sagittae.
It is also possible that annulus formation is the result of an energy shift from
somatic growth to reproductive output during spawning season, as observed in
Scomberomorus cavalla (Cuvier) (King Mackerel) by de L. Sturm and Salter
(1989). This hypothesis could be evaluated by maintaining Cobia in captivity yearround
to determine whether formation of opaque zones occurs in otoliths when fish
are prevented from participating in migrations but remain repro ductively active.
We obtained all specimens collected for the present study by sampling recreational-
angler catches during May–July. Our overall sample was truncated
by a minimum-fork-length harvest regulation (838 mm) enforced by the South
Carolina Department of Natural Resources. As a result, age-1 and age-2 fish
were underrepresented in our study. In addition, all samples were collected using
hook-and-line gear, which has been shown to have the potential to create a size
bias based on hook size or angler activity (Ralston 1990). This gear-selectivity
bias can be problematic because it tends to create a skewed representation of
the age structure of the stock, which can misinform assessments and impact the
estimated status of a fish stock (Cowan 2011). We recognize that our initial estimates
of the von Bertalanffy growth parameters may have been skewed by the
lack of data for age-1 fish, the truncation of age-2 data (Fig. 4A), and by gear
Southeastern Naturalist
725
C.A. Kalinowsky, M.C. Curran, and J.W. Smith
2016 Vol. 15, No. 4
selectivity. The lack of age-1 and some age-2 fish skewed our initial estimates
of K and t0. The variability in estimates of k and t0 between the present study and
studies from other regions (Table 3) is most likely the result of these missing age
classes. To overcome this missing component, we tried using data on younger
fish from nearby (NC) as well as analytical techniques (Diaz et al. 2004) to try
and adjust for those ages we lacked.
The Diaz-adjusted von Bertalanffy method has been effective in overcoming
truncated age data for parameter estimates for other fisheries (Diaz et al. 2004),
such as Red Snapper (SEDAR 2013c). In the current study, this method helped
adjust the growth curve toward 0, compensating for the missing age-1 fish and providing
more reasonable estimates for both K and t0. We assumed that t0 should be
similar for both sexes and that estimates of t0 for males were probably skewed by
the paucity of age-2 males (n = 1). Using a fixed t0 for males with the Diaz-adjusted
method lowered the slope of the growth curve and provided a more realistic K value
for males; nevertheless, this approach may have underestimated L∞ for males (Fig.
4C). For comparative purposes, age-1 data (n = 9) from an earlier study of Cobia
from North Carolina (Smith 1995) were used to generate additional sex-specific
growth curves (Fig. 4D). This method provided the most realistic estimates for all
parameters (Fig. 4D). Based on the parameters estimated by all 4 von Bertalanffy
growth models, female Cobia in South Carolina achieved greater theoretical asymptotic
length than did males (Fig. 4).
Currently, coastal development along the southeastern coast of the US is occurring
at accelerated rates. Population densities grew from 55.5 persons per km2 in
1980 to a projected 94.1 persons per km2 in 2008 (Crossett et al. 2004). Williams
et al. (2008) found that biomass of targeted fish species was negatively correlated
to the local human population in the Hawaiian Islands, indicating increased fishing
pressure in areas with increased coastal populations. We presume fishing pressure
on Cobia will increase; thus, a better understanding of the species’ life-history is
necessary for the effective management of the fishery. At present, few definitive
statements can be made regarding Atlantic Coast stock status for Cobia, mainly
because of the lack of comprehensive life-history information for the species (Williams
2001). The most current assessment indicates that Cobia stocks (both Gulf
and Atlantic) are not overfished (SEDAR 2013a, b). As a result of that assessment,
annual catch limits were established for both stocks in order to comply with the
Magnuson–Stevens Reauthorization Act of 2006. These measures are a step in
the right direction, but additional work is needed to further improve assessment
estimates and adjust catch limits (SEDAR 2013a, b).
Continued collection of data to improve regional stock assessments is needed for
populations of Cobia in the Western Central Atlantic and Gulf of Mexico to better
elucidate the overall stock status of Cobia. The life-history characteristics of Cobia
as defined by this study provide managers with additional age and growth information
that help to effectively manage the species. For example, the data collected
during this study was used during the most recent assessment (SEDAR 2013b)
on Cobia and provided managers with growth parameters from a region that had
Southeastern Naturalist
C.A. Kalinowsky, M.C. Curran, and J.W. Smith
2016 Vol. 15, No. 4
726
not been studied before. Future research should focus on tagging Cobia to better
describe migratory behavior and identify specific overwintering locations, and on
sampling age 0–1 fish to increase the accuracy of t0 estimates.
Acknowledgments
We thank S. Lynn and S. Kalinowsky for all of their assistance in the field. We are
grateful to the staff of the South Carolina Department of Natural Resources (SCDNR):
M. Denson, W. Jenkins, K. Brenkert, J. Yost, C. Taylor, and all other staff, interns, and
volunteers of the SCDNR Marine Resources Division (MRD) that participated in this
effort. We benefitted from the assistance of J. Smith, D. Vaughan (NMFS-Beaufort
Laboratory), E. Robillard, and Sonny Emmert (Georgia DNR-CRD). We thank the marina
operators and tournament officials who allowed us to sample at their facilities and
events including: Lemon Island Marina, Bryans Seafood, Hilton Head Food and Beverage
Tournament, Fillin’ Station Tournament, Lowcountry Marine Tournament, and Hoots
Tournament. We also appreciate the recreational anglers that allowed us to sample their
catches and the Hilton Head Sportfishing Club for the continued support throughout this
effort. We are grateful to M. Sherman for extensive editing assistance funded by a Department
of Education Title VII grant awarded to M.C. Curran (# P382G090003) and to
the NIH SCORE program for funding assistance. Finally, we thank Capt. J. Clark, Capt.
B. Parker, Capt. J. Deloach, Capt. M. Upchurch, Capt. J. Walker, and all other charterboat
captains that donated their time and effort to help make this project a success. This
publication is also listed as Contribution Number 1838 of the Belle W. Baruch Institute
for Marine and Coastal Science.
Literature Cited
Burns K.M., C. Neidig, J. Lotz, and R. Overstreet. 1998. Cobia (Rachycentron canadum)
stock-assessment study in the Gulf of Mexico and in the South Atlantic. Technical Report
No. 571. Mote Marine Laboratory, St. Petersburg, FL. 108 pp.
Chiang, W.C., C.L. Sun, S.Z. Yeh, and W.C. Su. 2004. Age and growth of Sailfish (Istiophorus
platypterus) in waters off eastern Taiwan. Fishery Bulletin 102:251–263.
Cowan, J.H., Jr. 2011. Red Snapper in the Gulf of Mexico and US South Atlantic: Data,
doubt, and debate. Fisheries 36(7):319–331.
Crossett, K.M., T.J. Culliton, P.C. Wiley, and T.R. Goodspeed. 2004. Population trends
along the coastal United States: 1980–2008. National Oceanic and Atmospheric Administration,
National Ocean Service, Management and Budget Office, Special Projects.
Available online at (http://stics.noaa.gov). Accessed 1 November 2008.
de L. Sturm, M.G., and P. Salter. 1989. Age, growth, and reproduction of the King Mackerel,
Scomberomorus cavalla (Cuvier), in Trinidad waters. Fishery Bulletin 88:361–370.
Diaz, G.A., C.E. Porch, and M. Ortiz. 2004. Growth models for Red Snapper in US Gulf
of Mexico waters estimated from landings with minimum-size-limit restrictions. SEDAR7-
AW-01. National Marine Fisheries Service, Miami, FL.
Franks, J.S., M.H. Zuber, and T.D. McIlwain. 1991. Trends in seasonal movements of
Cobia, Rachycentron canadum, tagged and released in the northern Gulf of Mexico.
Journal of the Mississippi Academy of Science 36(1):55.
Franks, J.S., J.R. Warren, and M.V. Buchanan. 1999. Age and growth of Cobia, Rachycentron
canadum, from the northeastern Gulf of Mexico. Fishery Bulletin 97:459 –471.
Southeastern Naturalist
727
C.A. Kalinowsky, M.C. Curran, and J.W. Smith
2016 Vol. 15, No. 4
Haddon, M. 2001. Modelling and Quantitative Methods in Fisheries. CRC Press, Boca
Raton, Florida. 450 pp.
Hammond, D.L. 2001. Status of the South Carolina fisheries for Cobia. Technical Report.
South Carolina Department of Natural Resources, Charleston, SC. 21 pp.
Herre, A.W. 1953. Check-list of Philippine fishes. Research Report. US Department of Interior,
Fish and Wildlife Service, Washington, DC. 977 pp.
Howse, H.D., R.M. Overstreet, W.E. Hawkins, and J.S. Franks. 1992. Ubiquitous perivenous
smooth-muscle cords in viscera of the teleost Rachycentron canadum, with special
interest on liver. Journal of Morphology 212:175–189.
Hrincevich, A.W. 1993. Analysis of Cobia Rachycentron canadum population structure in
the northern Gulf of Mexico using mitochondrial DNA. M.Sc. Thesis. University of
Southern Mississippi, Hattiesburg, MS. 91pp.
Kimura, D.K. 1980. Likelihood methods for the von Bertalanffy growth curve. Fishery
Bulletin 77:765–776.
Lefebvre, L.S. 2009. Inshore spawning of Cobia (Rachycentron canadum) in South Carolina.
M.Sc. Thesis. College of Charleston, Charleston, SC. 69 pp.
Ralston, S. 1990. Size selection of snappers (Lutjanidae) by hook-and-line gear. Canadian
Journal of Fisheries and Aquatic Sciences 47:696–700.
Richards, C.E. 1967. Age, growth, and fecundity of the Cobia, Rachycentron canadum,
from Chesapeake Bay and adjacent mid-Atlantic waters. Transactions of the American
Fisheries Society 96.3:343–350.
Ricker, W.E. 1975. Computation and interpretation of biological statistics of fish populations.
Bulletin of the Fisheries Research Board of Canada 191:1 –382.
Robins, C.R., and G.C. Ray. 1986. A Field Guide to Atlantic Coast Fishes of North America.
Houghton Mifflin Company, Boston, MA. 354 pp.
SAS Institute Inc. 1999. SAS/STAT User’s Guide, Version 8. SAS Institute Inc., NC.
SEDAR. 2013a. SEDAR 28 - South Atlantic Cobia Stock-Assessment Report. North
Charleston SC. 420 pp.
SEDAR. 2013b. SEDAR 28 - Gulf of Mexico Cobia Stock-Assessment Report. North
Charleston SC. 616 pp.
SEDAR. 2013c. SEDAR 31 - Gulf of Mexico Red Snapper Stock Assessment Report. SEDAR,
North Charleston SC. 1103 pp.
Shaffer, R.V., and E.L. Nakamura. 1989. Synopsis of biological data on the Cobia, Rachycentron
canadum (Pisces: Rachycentridae). NOAA Technical Report NMFS 82, FAO
Fisheries Synopsis 153. US Department of Commerce, Silver Sprin g, MD. 26 pp.
Smith, J.W. 1995. Life history of Cobia, Rachycentron canadum (Osteichthyes: Rachycentridae),
in North Carolina waters. Brimleyana 23:1–23.
Sun, C.L., S.P. Wang, and S.Z. Yeh. 2002. Age and growth of Swordfish (Xiphias gladius
L.) in the waters around Taiwan determined from anal-fin rays. Fishery Bulletin
100:822–835.
Thompson, B.A., C.A. Wilson, J.H. Render, and M. Beasley. 1991. Age, growth, and reproductive
biology of greater Amberjack and Cobia from Louisiana waters: One-year report
to US Department of Commerce, NOAA, National Marine Fisheries Service, Center for
Wetland Resources, Louisiana State University, Baton Rouge. LA. Cooperative Agreement
NA90AAH-MF089. 55 pp.
Tserpes, G., and N. Tsimenides. 1995. Determination of age and growth of Swordfish,
Xiphias gladius L., 1758, in the eastern Mediterranean using anal-fin spines. Fishery
Bulletin 93:594–602.
Southeastern Naturalist
C.A. Kalinowsky, M.C. Curran, and J.W. Smith
2016 Vol. 15, No. 4
728
Williams, E.H. 2001. Assessment of Cobia, Rachycentron canadum, in the waters of the
US Gulf of Mexico. NOAA Technical Memorandum. NMFS-SEFSC-469. Silver Spring,
MD. 55 pp.
Williams, I.D., W.J. Walsh, R.E. Schroeder, A.M. Friedlander, B.I. Richards, and K.A.
Stamoulis. 2008. Assessing the importance of fishing impacts on Hawaiian coral-reef
fish assemblages along regional-scale human-population gradients. Environmental Conservation
35:261–272.