448 
Fishery Bulletin 11 5(4) 
dination space. Average proportion of taxa per station 
drove assemblage structure defined by the beam trawl 
samples (Fig. 4) because the beam trawl caught a small¬ 
er size range of individuals. Average individual weights 
by taxon per station were more important to assemblage 
structure for bottom trawl samples (Fig. 5). 
Sampling season and variation in mesh sizes of nets 
may account for some of the differences in the compari¬ 
sons of assemblages for the 2 gear types. Beam trawl 
sampling occurred earlier in the spring than bottom 
trawl sampling, and the beam trawl had a smaller 
mesh. Additionally, variability in sampling time of day 
(lack of day samples for beam trawl collections) may 
have influenced our analyses. Other research on the 
northeast U.S. shelf has shown that gear type and 
mesh size of nets influence species and size in compo¬ 
sition of catch (Vasslides and Able, 2008; Slacum et al., 
2010; Malek et al., 2014). Beam trawls often catch more 
demersal taxa and otter trawls catch more pelagic taxa 
(Vasslides and Able, 2008; Malek et al., 2014). Malek et 
al. (2014) also found that a beam trawl caught smaller 
individuals than an otter trawl equipped with a net 
of the same mesh size. Assessments of WEAs would 
clearly benefit from the use of gears, such as beam 
trawls, that collect smaller individuals of both fish and 
macro-invertebrates, allowing a more comprehensive 
understanding of the potential impact of developing 
wind farms, because the combination of beam trawls 
with bottom trawls provides a more complete view of 
demersal communities than bottom trawls alone. 
Fish and macro-invertebrate assemblages varied 
spatially on the northeast U.S. shelf for both gear 
types, and latitude and longitude were important ex¬ 
planatory variables. The VA WEA is the farthest dis¬ 
tance from the other WEAs and had the least overlap 
in assemblage composition for the beam trawl stations 
(Fig. 4). The NJ and NY WEAs were the closest to each 
other, and had the most overlap in assemblages for the 
bottom trawl stations (Fig. 4), and both also overlapped 
with portions of RIMA-MA for beam trawl stations 
(Fig. 3). Regional variation in assemblage structure 
on the northeast U.S. shelf, following a south to north 
gradient, has been described for both fish (Gabriel, 
1992; Lucey and Nye, 2010) and macro-invertebrates 
(Wigely and Theroux, 1981; Theroux and Wigely, 1998; 
Hale, 2010). Consequently, impacts of WEAs spread 
along the shelf may become additive for assemblages 
that span large distances (i.e., 50-100 km). Therefore, 
impact assessments need to take a more holistic eco¬ 
system-scale approach. 
Assemblage structure varied within WEAs, and may 
be due to the effect of habitat conditions on species 
distributions. The explanatory habitat variables, sedi¬ 
ment, depth, and bottom water temperature, correlat¬ 
ed with assemblage structure in the beam trawl and 
bottom trawl collections (Fig. 6). The lack of daytime 
beam trawl sampling may have influenced our results 
because some species have been found to exhibit diel 
patterns of microhabitat use on the northeast U.S. 
shelf (Diaz et al., 2003); consequently additional round- 
the-clock sampling may lead to the discovery of differ¬ 
ent habitat relationships than those we report here. 
Nevertheless, taxon-specific relationships with habi¬ 
tat on the shelf have previously been shown for fish 
and macro-invertebrates and were most closely associ¬ 
ated with sediment characteristics (Wigely and Ther¬ 
oux, 1981; Theroux and Wigely, 1998; Methratta and 
Link, 2006, 2007), depth (in both the cross-shelf and 
shoal formations) (Viscido et al., 1997; Steves et al., 
1999; Methratta and Link, 2007; Vasslides and Able, 
2008; Slacum et al., 2010), and bottom water tempera¬ 
ture, particularly with seasonal temperature changes 
(Steves et al., 1999; Malek et al., 2014). Hale (2010), in 
examining estuarine and near shore sample locations, 
also showed that salinity was related to macro-inver¬ 
tebrate assemblages. The RIMA-MA WEA covered the 
largest area and had the most diverse assemblages for 
both beam and bottom trawl collections (Figs. 4 and 5). 
This finding may be related to greater heterogeneity in 
habitat types within this combined WEA. Impact as¬ 
sessments within and among WEAs need to take into 
account habitat variability. 
Our study provides a “snapshot” of the springtime 
assemblages for the northeast U.S. shelf. Seasonality, 
as defined by day of year, explained a small proportion 
(<4%) of the variability in fish and macro-invertebrate 
assemblage structure for both the beam and bottom 
trawl collections. Other research on the northeast U.S. 
shelf has documented the existence of seasonal assem¬ 
blages. Steves et al. (1999) identified 3 seasonal assem¬ 
blages for recently settled juvenile fish: winter-spring, 
summer, and fall. Malek et al. (2014) identified sum¬ 
mer and fall assemblages from beam trawl and otter 
trawl collections in Rhode Island Sound that remained 
stable across years. Therefore yearlong temporal sam¬ 
pling of WEAs will be needed to assess the entire as¬ 
semblage structure, especially considering the complex 
life history (e.g., multiple life stages with various habi¬ 
tat needs) of many of the shelf species of both fish and 
macro-invertebrates. 
Overall, our results and previous research indicate 
that effects of WEAs will need to be critically assessed 
for WEAs on an individual basis. Additionally, effects 
should be evaluated across multiple spatial and tempo¬ 
ral scales to determine population-level effects on resi¬ 
dent fish and macro-invertebrates (Bergstrom et al., 
2014; Lindeboom et al., 2015). Many species use large 
areas during their life span (e.g., egg, larval, juvenile, 
adult stages), often make long-distance seasonal mi¬ 
grations (Secor, 2015), and therefore small-scale effects 
(on the scale of individual WEAs) may have additive 
effects (Bergstrom et al., 2014; Dai et al., 2015) on 
populations of commercially and ecologically important 
fish and macro-invertebrates. 
Acknowledgments 
Partial funding for this project was provided by the 
BOEM-NOAA Interagency Agreement #M13PG00019. 
