234 
Fishery Bulletin 119(4) 
in the LLS compared with the rate of capture in the BTS, 
it would be unclear whether it is due to low selectivity in 
the longline gear or simply due to the absence (i.e., low 
availability) of these fish. To address this question, we 
used white hake as a control for the selectivity of the long- 
line gear. Atlantic cod and white hake, the latter of which 
has an accepted dome-shaped selectivity for the BTS, are 
generally similar gadoids in size (both can exceed 100 cm 
TL), prey types, and prey size (Scharf et al., 2000). There- 
fore, we would expect both species to have comparable 
probability of capture with the LLS gear. In other words, 
if large white hake (=90 cm TL) are caught in the LLS, we 
would expect large cod to be caught as well, if these fish 
are present. Conversely, if large white hake are caught in 
the LLS but large cod are not, we would deduce that large 
cod are not present in the area. This approach offers an 
avenue to examine whether data from the LLS supports 
the presence of large cod in the rough-bottom habitat of 
the GOM. 
The objectives of this study were to evaluate the relative 
size distributions of Atlantic cod and white hake in the LLS 
and BTS and to test the hypothesis that availability by size 
is different between the predominantly smooth-bottom 
habitats sampled with bottom trawl gear and the rough- 
bottom habitats sampled with longline gear. Specifically, 
we will focus on the upper end of the size distributions of 
the LLS and BTS catches to explore whether they differ 
for each species. Depending on the results, as explained 
previously, our findings may indicate the need for recon- 
sideration of the selectivity curve used for the BTS in the 
GOM Atlantic cod stock assessment. In addition, results of 
our analysis can also validate the current BTS selectivity 
assumed for the white hake assessment. 
Materials and methods 
The NEFSC has been conducting the BTS in the fall since 
1963 and in the spring since 1968 (Azarovitz, 1981). This 
survey follows a stratified random sampling design, with 
stratification by depth and region, for selection of sam- 
pling stations in federal waters on the continental shelf 
from Cape Hatteras to the Scotian Shelf (Stauffer, 2004; 
Politis et al.°). Stations are randomly selected within each 
stratum prior to each survey cruise, with the number of 
stations selected per stratum being proportional to the 
area of the stratum, typically with a minimum of 3 sta- 
tions per stratum (Politis et al.°). Although several factors 
are considered before a tow begins (e.g., presence of ship- 
wrecks or fixed gear and strength of bottom current), bot- 
tom type does not specifically affect the determination of a 
tow location. Since 2009, the NEFSC has used the NOAA 
Ship Henry B. Bigelow as the primary research vessel for 
the BTS. The NOAA Ship Pisces was used in the fall of 
2017; however, because the 2 vessels are the same size and 
class it is assumed that changes in vessel did not affect 
selectivity. Each vessel used a 4-seam, 3-bridle bottom 
trawl with a rockhopper sweep (Politis et al.°) that can be 
used in a variety of habitat types. At each station, data on 
the abundance, biomass, and size distribution of fish were 
collected for each species. Fish TL was measured (to the 
nearest 0.5 cm) by using digital fish measuring boards. 
The LLS, which is an industry-based survey imple- 
mented by the NEFSC (McElroy et al., 2019), has occurred 
concurrently with the spring and fall BTS since 2014. 
The further stratification of sampling by bottom type for 
the LLS is based on rugosity. Bottom type was defined as 
either smooth or rough, with rough-bottom habitats hav- 
ing a terrain ruggedness index value (Riley et al., 1999) 
greater than the 70th percentile within each stratum 
(McElroy et al., 2019). With the participation of 2 com- 
mercial fishing vessels, the sampling of the LLS occurs 
between 42.0°N and 43.5°N, an area that encompasses 
all or portions of 6 BTS strata (Fig. 2). Sampling density 
is higher in the LLS than in the BTS, and stations are 
more heavily allocated to rough-bottom substrata in the 
LLS (Politis et al.°; McElroy et al., 2019). In the current 
allocation of stations sampled in the LLS, 84% of stations 
are in rough-bottom strata, and the sampling in the BTS 
when stratified post hoc by using the LLS strata includes 
20-35% of stations in the rough-bottom strata per year 
(Suppl. Table) (online only). 
Tub-trawl longlines are used for the LLS with gear simi- 
lar to those used commercially for groundfish species, such 
as cod and white hake, in the region. The longline gear 
is deployed with 1000 12/0 semicircle E-Z baiter hooks’ 
(O. Mustad and Son AS, Gjgvik, Norway) on 38-cm gan- 
gions that are baited with squid on a 1.85-km (1-nmi) 
mainline and soaked for 2 h across the slack tide (McElroy 
et al., 2019). The hook size and soak time were chosen to be 
relatively comparable with some commercial groundfish 
operations, and the soak time was standardized across the 
slack tide to get a portion of both sides of a tide. The main- 
line length was chosen to be consistent with the mean tow 
distance of the BTS. At each station, data on the abun- 
dance, biomass, and size distribution of fish were collected 
for each species. Fish TL was measured (to the nearest 
0.5 cm) by using digital fish measuring boards. 
Length frequencies of cod and white hake caught in the 
LLS and BTS were generated by using TL by season for 
the period 2014—2018. For the BTS, data from all stations 
meeting gear and operational acceptance criteria (Politis 
et al.°) were included from stock assessment strata 
sets for each species: offshore strata 26-30 and 36—40 for 
the GOM cod stock (NEFSC, 2013a) and offshore strata 
21-30 and 36-40 for the white hake stock (NEFSC, 2013b) 
(Fig. 2). Use of stock assessment strata for the BTS 
reflects the intent of the analysis to relate the findings 
about availability of cod and white hake in the rough- 
bottom habitat of the GOM based on data from the LLS to 
the respective stocks as a whole, rather than simply com- 
paring BTS and LLS results in the overlapping area. For 
the LLS, data from all stations that met gear and opera- 
tional acceptance criteria (McElroy et al., 2019) were 
7 Mention of trade names or commercial companies is for identi- 
fication purposes only and does not imply endorsement by the 
National Marine Fisheries Service, NOAA. 
