Fairchild et al. : Spring feeding of Anarhichas lupus on Stellwagen Bank, Massachusetts 
193 
by an overdose of tricaine methanesulfonate (MS-222), 
and then heads, gonads, and entire gastrointestinal 
tracts were removed, bagged separately, and frozen un- 
til further analyses. 
Age 
Otoliths were processed at the Massachusetts Division 
of Marine Fisheries Age and Growth Laboratory, locat- 
ed at the Annisquam River Marine Fisheries Station 
in Gloucester, Massachusetts. The sagittal otoliths re- 
moved from each wolffish head were cleaned, embedded 
in 2 part epoxy, and sectioned transversely through the 
core. A low-speed Isomet 7 saw (Buehler, Lake Bluff, IL), 
with 2 diamond blades separated by a 0.4-mm spacer, 
was used to take sections. Otolith sections were affixed 
to microscope slides with Flo-Texx mounting medium 
(Thermo Fisher Scientific Inc., Waltham, MA) and then 
viewed through a compound microscope at 100-400x 
magnification for the enumeration of annuli. If the an- 
nuli were not clearly visible, the section was polished 
on a Buehler Ecomet 3000 variable-speed grinder-pol- 
isher until it was thin enough for easy enumeration. 
Annuli were defined as the thin hyaline zones described 
by Jonsson (1982) and were counted along the ventral 
side of the sulcal groove. Age was determined by 3 in- 
dependent readers. When discrepancies occurred, all 3 
readers reviewed the otoliths together and ages were 
decided by the majority. Coefficient of variation (Chang, 
1982) was calculated between each reader combination. 
A birth date of 1 January was used. 
Total mortality 
From the age frequency of the pooled samples (n= 303), 
Z was estimated by the method of Chapman and Rob- 
son (1960) with bias correction (Seber, 1982) for Z and 
a correction for over-dispersion for the standard error 
of Z (Smith et al., 2012). According to Murphy (1997) 
and Smith et al. (2012), the Chapman-Robson estima- 
tor performs better and is less biased than other avail- 
able estimators (e.g. linear regression). Calculations 
were performed with R statistical software, vers. 2.15.1 
(R Core Team, 2012) with the function agesurv in the 
R package fishmethods, vers. 1.3-0 (Nelson, 2012). The 
starting point was the age of full recruitment (age at 
maximum catch) plus 1 year (Smith et al., 2012). Sen- 
sitivity of the estimate to the starting point was exam- 
ined by running the same analysis but starting 1 year 
before and 1 year after the initial starting point. 
Sex and reproductive maturity 
Whole, paired gonads were thawed, identified as testes 
or ovaries, and weighed to the nearest 0.1 g. For fe- 
males, smallest and largest oocytes were measured (di- 
7 Mention of trade names or commercial companies is for iden- 
tification purposes only and does not imply endorsement by 
the National Marine Fisheries Service, NOAA. 
ameter to the nearest 0.1 mm), overall color was record- 
ed, and, if possible, gross morphology of oocyte stages 
was noted. Maturity guidelines published by Temple- 
man (1986a) and Gunnarsson et al. (2006) and based 
on egg size and color were used to classify maturity. 
In mature prespawning Atlantic wolffish, 3 generations 
of oocytes are present in the ovary: primary oocytes, 
which will advance to the cortical alveolus (CA) stage 
in the following year (<0.5 mm, whitish); oocytes in the 
CA stage to be spawned next year (0.5-1. 8 mm, yellow 
and orange); and oocytes to be spawned in the current 
year (spawning stage 3; 2. 5-4. 8 mm, deeper yellow and 
orange; Gunnarsson et al., 2006). Atlantic wolffish with 
oocytes in the CA stage (>0.5 mm) are deemed mature 
(Gunnarsson et al., 2006). To assess reproductive state, 
the gonadosomatic index (GSI) was calculated as GSI 
= W Gonads / (W Body ~ W Gonads^- 
Feeding ecology 
Gastrointestinal tracts were thawed, weighed to the 
nearest 0.1 g, and dissected, and all prey items were 
identified to the lowest possible taxon and weighed as 
a group to the nearest 0.1 g. Frequency of occurrence, 
volume of dietary items, feeding index, and Fulton’s 
condition factor ( K) were calculated with the following 
equations 
Frequency of occurrence = 
[number of stomachs containing prey item i 
/ total number of examined stomachs] x 100. 
Volume by species = [W prey item i / ^stomach) * 100. 
Feeding index — [Wtotal stomach contents I Wfishl x 100. 
K - W (g )/TL 3 (mm) x 105. 
Because Atlantic wolffish crush their prey and because 
of the prevalence of hard-shelled invertebrates like 
mollusks, crustaceans, and echinoderms, no attempts 
were made to quantify numerical abundance of prey. 
Statistical analyses 
A chi-square test of independence was used to deter- 
mine whether the sex ratio deviated from a 1:1 female- 
to-male ratio. Spearman rank order correlation tests 
were used to determine relationships between gonad 
size, GSI, and K and sex, age, TL, and W of Atlantic 
wolffish, and Mann-Whitney rank sum tests were used 
to compare TL, W, age, and K between male and female 
wolffish with SigmaPlot (vers. 11.0, Systat Sofware 
Inc., San Jose, CA). 
Results 
In the study area, 53 tows were completed, resulting 
in the capture of 395 Atlantic wolffish (Table 1, Fig. 
1). The first 4 days of sampling yielded 304 wolffish, 
all of which, except for 1, were euthanized for life his- 
