286 
Fishery Bulletin 115(3) 
ded in paraffin. Tissues were cross sectioned at 4 pm 
and stained with hematoxylin and eosin. Microscopic 
inspection of testicular and ovarian tissues and evalu- 
ation of spermatogenesis and oogenesis followed the 
method of Brown-Peterson et al. (2011). Reproductive 
phases were assigned to males and females (Brown-Pe- 
terson et al., 2011), and included immature, developing 
(with the early developing subphase), spawning capa- 
ble (with the actively spawning sub-phase for females), 
regressing, and regenerating. Additionally, broad repro- 
ductive descriptors of males and females were defined 
as reproductively active (including the developing, 
spawning capable, and actively spawning phases) or 
reproductively inactive (including the immature, early 
developing, regressing, and regenerating phases). 
Females assessed macroscopically to be in the ac- 
tively spawning subphase were sampled for fecundity 
by placing a weighed portion (0.01 g) of the ovary into 
Gilson’s fixative or 10% neutral buffered formalin. 
Ovarian tissue fixed in Gilson’s fixative remained in 
solution for a minimum of 3 months prior to analy- 
sis; tissue fixed in 10% formalin remained in solution 
1 month before analysis. Tissues were rinsed overnight 
in running water and oocytes remaining attached to 
ovarian tissue were teased apart. The volumetric meth- 
od (Bagenal and Braum, 1971) was used to determine 
BF (number of eggs), and 6 replicate samples of all oo- 
cytes >500 pm, representing oocytes undergoing oocyte 
maturation (OM), were counted. To determine the cor- 
rect size of oocytes to count for fecundity determina- 
tions, all oocytes >80 pm from an actively spawning 
fish were counted and measured in a 1-mL subsample. 
Oocyte size-frequency graphs were constructed, and 
only oocytes >500 pm that presented a distinct mode at 
the end of the distribution were counted. This largest 
batch of oocytes was verified histologically to be either 
hydrated or undergoing OM for each sample analyzed. 
Additionally, the oocyte size-frequency distribution of 
secondary growth oocytes (>80 pm) of fish in the ac- 
tively spawning sub-phase at the beginning (October) 
and end (March) of the reproductive season was plot- 
ted to identify the fecundity pattern (determinate or 
indeterminate) of Gulf menhaden, following procedures 
described by Lowerre-Barbieri et al. (2011). 
Spawning frequency of females was determined by 
calculating the percentage of females in the spawning 
capable phase (including the actively spawning sub- 
phase) with 1) <24 h postovulatory follicle complex 
(POF) or 2) oocytes undergoing OM after procedures as 
outlined in Brown-Peterson and Warren (2001). Spawn- 
ing frequency data were used in combination with BF 
data to estimate total annual fecundity. 
Using fecundity data in the stock assessment 
The Beaufort Assessment Model (BAM) is the primary 
assessment model used for the federal stock assess- 
ment of Gulf menhaden (Williams and Shertzer, 2015). 
Data inputs include a suite of fishery-independent and 
fishery-dependent data, and the model is age struc- 
tured. Aging error was included in the model, as in 
the “base” formulation of the BAM during the last 
benchmark assessment (see SEDAR 2 for the aging er- 
ror matrix). Natural mortality was assumed to be time 
invariant, but age varying. The sex ratio was fixed at 
1:1 and the parameters of the von Bertalanffy growth 
function were estimated internally and constant. Fish 
were assumed to spawn on 1 January, and therefore all 
reproductive values in the stock assessment correspond 
to 1 January. The maturity ogive was fixed such that 
0% of individuals were mature at age 0 and age 1 and 
100% were mature at age 2. Mean lengths at age on 1 
January were used to calculate mean fecundity at age 
on 1 January. Recruitment was modeled with a Bever- 
ton-Holt stock recruitment curve and steepness was 
fixed at 0.75. Selectivity for the commercial reduction 
fishery was assumed to be dome shaped and only age- 
1 selectivity was estimated. Constant catchability was 
estimated for each of the 2 fishery-independent gears 
(gill and seine nets) used to sample the stock. Gillnet 
sampling targets age-1 to 4+ fish, whereas seine net 
sampling targets predominately age-0 fish. 
We compared the estimates from the “base” formula- 
tion of the BAM, that incorporates age-specific fecundi- 
ty estimates from Lewis and Rothymayr (1981) and the 
maturity schedule described above, with BAM-derived 
estimates based on alternative formulations from re- 
productive dynamics described in this work. We incor- 
porated 2 changes to the input variables in the BAM 
model: the magnitude of individual egg production 
and the maturity schedule. We used “high” and “low” 
estimated annual fecundity that were based on the 
mean length-specific egg production of each age class, 
given length-at-age on 1 January, and the minimum 
and maximum estimated number of annual spawns 
for each age class. The high annual fecundity scenario 
assumes more frequent spawning for each individual 
relative to the low annual fecundity scenario. We also 
altered the age-specific maturity schedule based on the 
length-specific maturity relationship we developed. We 
evaluated the temporal trends in the magnitude of egg 
production of the stock and the maximum proportion of 
spawners-per-recruit of the stock for each scenario. We 
also evaluated the year-specific percent difference in 
number at age and instantaneous annual fishing mor- 
tality ( F ) at age between the base model used in the 
2012 stock assessment and the base model used in the 
2013 stock assessment, using the logistic relationship 
described in this work. 
Comparisons of historical data 
To assess historic changes in the Gulf menhaden stock 
in the northern Gulf of Mexico (Mississippi and Loui- 
siana) we used data (NMFS 6 ) obtained from the com- 
NMFS (National Marine Fisheries Service). 2015. Unpubl. 
data. Gulf menhaden biostatistical database. [Available 
from Beaufort Laboratory, Southeast Fish. Sci. Center., Natl. 
Mar. Fish. Serv., 101 Fivers Island Rd., Beaufort, NC 28516.] 
