26 
Fishery Bulletin 99(1 ) 
It is reasonable to assume that a female will not always 
spawn the same number of oocytes during each spawning 
event and that this variation in batch size may not be 
related to body size, as indicated for fecundity estimates 
with the FOM method. Furthermore, although only large 
oocytes were counted with the GF and NBF methods, not 
all females with oocytes >700 pm underwent FOM, as 
determined by histological inspection. Because histologi- 
cal inspection is not always practical when fecundity esti- 
mates are taken, we feel our estimates should include all 
fish with oocytes >700 pm. Both the GF and NBF methods 
resulted in higher, but not significantly different, fecundi- 
ty estimates than those yielded by the histological meth- 
od, suggesting that the wide variation among individual 
fish obscures any meaningful difference among methods. 
We believe our most accurate fecundity estimates are 
based on the actual histological counts of oocytes undergo- 
ing FOM. Our approach is supported by Hunter and Mace- 
wicz’s (1985b) finding that oocytes undergoing FOM can 
be used for fecundity estimates in fish with rapid FOM 
when hydrated oocytes are unavailable. Although the ex- 
act time frame of FOM is unknown for cobia, we presume 
it is relatively rapid. For example, fish in the early stages 
of FOM were captured in the morning. Cobia are pre- 
sumed to spawn during the day, probably the late after- 
noon, on the basis of collections of fertilized eggs (Ditty 
and Shaw, 1992). Other multiple spawning fish from simi- 
lar latitudes (e.g. C. nebulosus [see Brown-Peterson et al., 
1988], black drum, Pogonias cromis [see Fitzhugh et al., 
1993], and C. undecimalis [see Taylor et al., 1998]) under- 
go FOM within 12 h. Several large scombrids also have 
rapid FOM (McPherson, 1993; Schaefer, 1996; Farley and 
Davis, 1998). From this evidence, we conclude that accu- 
rate batch fecundity estimates can be made for cobia by 
using oocytes from fish undergoing FOM. 
The estimated mean batch fecundity values from the 
present study (1.9 x 10 6 eggs with the GF method, 8.5 x 
10 5 with the NBF method, and 3.8 x 10 5 with the histo- 
logic method) are lower than previous mean estimates by 
Lotz et al. (1996) of 4.8 x 10 7 and Richards (1967) of 2-5 
x 10 6 eggs. Differences in methods no doubt explain the 
wide range in estimates. We used only oocytes >700 pm 
for fecundity estimates rather than all oocytes >550 pm 
used by Lotz et al. (1996) and Richards (1967). Our meth- 
ods ensured that only oocytes likely to undergo hydration 
within the following 24 h were included in fecundity es- 
timates. Lotz et al. (1996) suggested that their batch fe- 
cundity values may have been an overestimate because 
all the oocytes counted may not have been released dur- 
ing spawning. Richards (1967) probably also overestimat- 
ed the batch fecundity of cobia, although his estimates are 
close to ours obtained by using the GF method. 
Batch fecundity was not estimated for any species by us- 
ing direct histological counts of oocytes undergoing FOM; 
thus, it is difficult to compare our results with other pub- 
lished results. Even though the estimates appear low 
when compared with more traditional methods of estimat- 
ing fecundity, there is less variation in the counts. The 
relatively small sample size {n= 26) used in our study for 
the FOM method may have resulted in an underestima- 
tion of batch fecundity. Increasing the sample size from 
<30 to 298 fish resulted in an increase as great as 33% in 
batch fecundity estimates for Atlantic mackerel (Scomber 
scombrus, see Watson et al., 1992). Although our batch fe- 
cundity estimates for cobia are realistic first approxima- 
tions, additional samples are necessary to produce a more 
accurate mean estimate of batch fecundity, a crucial value 
for accurate spawning stock biomass assessments. In ad- 
dition, the large variations in batch fecundity among indi- 
vidual fish are probably a biologically accurate represen- 
tation of variations in batch size in this multiple spawning 
species. Thus, assigning a single value to the batch fecun- 
dity of cobia does not give a biologically accurate portrayal 
of spawning stock biomass. 
The mean relative fecundity of 29.1 ±4.8 to 53.1 ±9.4 
eggs/g ovary-free body weight calculated for cobia is low 
when compared with co-occurring inshore and estuarine 
fish in the region (Brown-Peterson et al., 1988; Fitzhugh 
et al., 1993). Cobia, like the co-occurring tripletail (Lo- 
botes surinamensis) , wahoo ( Acanthocybium solandri), 
common dolphinfish (Coryphaena hippurus), king mack- 
erel ( Scomberomorus caualla), and greater amberjack {Se- 
riola dumerili), is a large, subtropical pelagic fish and ex- 
hibits a very different life history than smaller nearshore 
and estuarine species. Fecundity data are available only 
for two of these co-occurring species: wahoo, with an esti- 
mated relative fecundity of 57.7 eggs/g (Brown-Peterson 
et al., 2000) and tripletail, with an estimated relative fe- 
cundity of 47.6 eggs/g (Brown-Peterson and Franks, in 
press). Values from both species compare favorably with 
our estimates for cobia. Other pelagic species for which 
relative batch fecundity values are available include At- 
lantic mackerel (55.5 eggs/g; Watson et al., 1992), south- 
ern bluefin tuna ( Thumnus maccoyii, 57 eggs/g; Farley and 
Davis, 1998) and yellowfin tuna (Thunnus albacares, 68 
eggs/g; Schaefer, 1996). When the relative fecundity of the 
cobia is compared with that of other species with similar 
habitats and life histories, our estimate appears within 
the range of reported values for other pelagic species. 
Our study represents the first report of spawning fre- 
quency for cobia. The FOM and the POF methods pro- 
duced estimates of spawning at five-day intervals for co- 
bia in the SEUS and NCGOM. However, these estimates 
are based on three months of data for a potential six- 
month spawning season and thus may not represent the 
spawning frequency throughout the entire reproductive 
period for each region. Regardless, this spawning frequen- 
cy is lower than that reported for other large pelagic spe- 
cies, such as narrow barred Spanish mackerel (S. com- 
merson, 2-3 d, McPherson, 1993), southern bluefin tuna 
(daily spawners, Farley and Davis, 1998), yellowfin tuna 
(1-2 d, Schaefer, 1996) and wahoo (2-6 d, Brown-Peterson 
et al., 2000), as well as for the smaller pelagic carangids 
(3 d, Clarke and Privitera, 1995), spotted seatrout (2-7 d, 
Brown-Peterson et al., 1988), common snook (1. 1-2.5 d, 
Taylor et al., 1998), and red drum (2-4 d, Wilson and 
Neiland, 1994). Perhaps, the longer intervals between 
spawnings for cobia may be due to the longer distances 
that cobia need to travel between feeding and spawning 
grounds in comparison with the distances traveled by the 
