Walsh and Guida: Fish and macro-invertebrate assemblages near wind energy areas 
439 
trawl (Kuipers, 1975) with a 0.63-cm-mesh net from 11 
March to 12 April 2014. This sampling was restricted 
to the time of night because the research vessel was 
used to conduct visual transect surveys of marine mam¬ 
mals and precluded any fishing during daylight hours. 
Bottom trawl collections were made with a 4-seam, 
3-bridle otter trawl net with a dimension of 400x12 cm 
and with a 2.5-em-mesh liner (Bonzek et al. 4 ; Politis 
et al. 3 ). The nets used by NEFSC and NEAMAP for 
bottom trawl surveys are not identical, and differences 
include the sweeps, headline floats, number of both 
top and belly panels, and the mesh size of some panels 
(Bonzek et al. 4 ; Politis et al. 3 ). Samples were collected 
both day and night from 2 to 20 April 2014 during the 
NEFSC survey and during the day from 7 to 20 May 
2014 during the NEAMAP survey. 
Samples were collected with the beam trawl within 
the BOEM-designated lease blocks of 5 WEAs: Virginia 
(VA, n=12); New Jersey (NJ, n= 13); New York (NY, n=10); 
Rhode Island-Massachusetts (RIMA); and Massachu¬ 
setts (MA). Data from 2 BOEM WEAs, RIMA and MA, 
were combined into a single WEA (RIMA-MA, n= 23) 
because of their close proximity to one another (Fig. 
1). Lease blocks were chosen haphazardly to best coin¬ 
cide with visual transect surveys of marine mammals 
that occurred during daylight hours, and preference 
was given to those lease blocks that had been sampled 
during previous surveys with various gears. Both the 
NEFSC and NEAMAP surveys use a stratified random 
sampling design. Therefore, samples from within a 20- 
km buffer around each WEA with the same depth range 
as that of beam trawl stations (20-60 m) were used for 
comparison with beam trawl collections. This procedure 
resulted in bottom trawl samples being available for 
3 WEAs: NJ (n=10); NY (n=6); and RIMA-MA (n=23). 
Unfortunately, no bottom trawl samples were available 
for comparison with the VA WEA because of a lack of 
sampling during the NEFSC survey and because the 
WEA was deeper than the NEAMAP survey area. 
Trawl samples were processed on board, and pro¬ 
cessing was similar for the surveys at the NJ, NY, and 
RIMA-MA WEAs. Samples were sorted to the lowest 
practicable taxon, which varied by survey (Suppl. Ta¬ 
ble). We compared taxa sampled with the beam trawl 
with taxa sampled with the bottom trawl (Table 1) 
and found that shrimp were identified to lower taxo¬ 
nomic levels in the beam trawl collections (e.g., cari- 
dean shrimp and the white shrimp, Penaeus setiferus; 
Suppl. Table) (online only) than in NEFSC bottom trawl 
collections (e.g., unclassified shrimp; Suppl. Table) (on¬ 
line only). These finer taxonomic levels combined for 
the comparisons by gear type, and those comparisons 
were made at the highest level of identification (e.g., 
unclassified shrimp). The total numbers and aggregate 
weights (measured in kilograms) were available for 
each taxon and each station. Percent frequency of oc¬ 
currence was calculated for each taxon, by gear type. 
The proportion of a taxon in the catch at a sampling 
station was calculated by dividing the aggregate weight 
of a taxon by the total weight of all taxa captured at 
a station. The estimated average individual weight for 
each taxon (measured in grams) was calculated by di¬ 
viding the total weight (measured in grams) by the to¬ 
tal count for each taxon. 
Environmental and habitat sampling was conducted 
concurrently with trawl sampling. Water temperature 
and salinity were measured at each trawl station by 
using either a Sea-Bird Scientific 5 SBE 19 SeaCAT 
conductivity, temperature, and depth profiler (Sea-Bird 
Scientific, Bellevue, WA) or Hydrolab MSS sonde (OTT 
Hydromet, Kempten, Germany). Bottom water tem¬ 
perature (measured in degrees Celsius) and salinity 
measurements were taken within 5 m of the bottom. 
Sediment samples were collected at beam trawl sta¬ 
tions by using a 0.04-m 2 or G.10-m 2 Young-modified Van 
Veen grab sampler. The Folk (1954) sediment classifi¬ 
cation system was used to classify beam trawl station 
sediments into the following categories: l=muddy sand; 
2=sand; 3=sand-slightly gravelly sand; 4=slightly 
gravelly sand; 5=slightly gravelly sand-gravelly sand; 
6=gravelly sand; V-gravelly sand-sandy gravel. 
Statistical analyses 
Several statistical analyses were undertaken in order 
to compare catches across gear types and across WEAs. 
Correspondence analysis (CA) was used to compare the 
fish and macro-invertebrate assemblages in relation 
to gear type and WEA and canonical correspondence 
analysis (CCA) was used to examine the relationship 
between assemblages and sampling program and en¬ 
vironmental variables. A Student’s Ntest was used to 
examine whether the beam and bottom trawls collected 
individuals of significantly different sizes. 
The software package “FactoMineK” (Le et al., 2008) 
in R, vers. 3.2.2 (R Core Team, 2015) was used to per¬ 
form CA on average proportions and individual weights 
by a taxon per station to describe proportion of fish 
and macro-invertebrates and individual size, by sta¬ 
tion. We conducted 3 analyses: beam and bottom trawl 
collections combined; beam trawl only collections; and 
bottom trawl only collections. Beam and bottom trawl 
stations were analyzed together to examine differences 
among the collections, by gear type. Assemblage com¬ 
position in relation to WEA was examined by analyz¬ 
ing beam trawl and bottom trawl samples separately. 
The inclusion of rare taxa in CA and CCA often leads 
to assemblage patterns similar to those in data where 
rare taxa have been removed (Marancik et al., 2005; 
Walsh et al. 2006) and rare taxa often increase species 
richness of some assemblages (Marancik et ah, 2005) 
or appear as outliers separate from larger assemblages 
(Walsh et al., 2006). To simplify our analyses, however, 
only taxa that had at least a 10% frequency of occur¬ 
rence were used in the analyses. Percent frequency of 
occurrence was based on all stations sampled for the 
5 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. 
