Walsh and Guida: Fish and macro-invertebrate assemblages near wind energy areas 
447 
0 
£ 
o 
-1 
5 
O -2 
O 
-3 
-4 
O'" 
O 
C0 
5 
o 
o 
-3 
NY 
NJ 
o VA 
-1 0 1 
CCA1 (9.7%) 
2 J 
-2 - 
B 
Depth <> 
u, \ 
Light 
Long x 
® ® ® ® C 
• * 
Temp 
* □ 
• RIMA-M/ 
□ 
O NY 
□ NJ 
- 2-1012 
CCA1 (18.2%) 
Figure 6 
Canonical correspondence analysis (CCA) or¬ 
dination plots, showing the scores and vari¬ 
ance explained for the first and second di¬ 
mensions (CCA1 and CCA2), of the sampling 
stations for each wind energy area, Rhode 
Island-Massachusetts and Massachusetts 
(RIMA-MA), New York (NY), New Jersey 
(NJ), and Virginia (VA), by using fish and 
macro-invertebrate taxa that occurred in at 
least 10% of the collections made in 2014 with 
a (A) beam trawl and (B) bottom trawl. The 
arrows depict the gradient of each explana¬ 
tory variable. 
trawl station groups and assemblages and 28.8% for the 
bottom trawl station groups and assemblages. Finally, 
covariables explained smaller proportions of the vari¬ 
ance, with day of year explaining 3.7% and 3.1% for the 
beam trawl and bottom trawl, respectively. For the bot¬ 
tom trawl analysis, program (i.e., NEFSC vs. NEAMAP) 
explained 12.0% of the variance. For beam trawl sta¬ 
tion groups and assemblages, latitude (F=2.57, df=l, 
P<0.006), longitude (F=3.94, df=l, P<0.001), sediment 
(17=4.10, df=l, P<0.001), and bottom water temperature 
(F=1.94, df=l, P<0.040) were significantly correlated for 
the beam trawl station groups and assemblages (Fig. 
6A). For the bottom trawl station groups and assem¬ 
blages, all 5 remaining variables—latitude (F=4.87, 
df=l, P < 0.001), longitude (F= 2.92, df=l, P<0.005), 
depth (F=2.25, df=l, 0 11), bottom water tempera¬ 
ture (F= 2.05, df=l, P<0.019), and light (F=3.82, df=l, 
P<0.001—were correlated (Fig. 6B). Thus, the station 
groups in ordination space were similar but not identi¬ 
cal among the CAs and CCAs in the comparisons of the 
separation of WEAs in Figure 6 with those in Figures 
4 and 5. 
The variable of location on the northeast U.S. shelf, 
defined by latitude and longitude, correlated most with 
station groups and assemblages for the beam and bot¬ 
tom trawl collections. Most VA WEA stations from the 
beam trawl collections separated from the northern 
WEAs along the first and second dimensions (Fig. 6A), 
and aligned with latitude and longitude, and explained 
45.1% of the environmental correlation. Bottom water 
temperature explained 8.8% of correlation and aligned 
opposite latitude and longitude (Fig. 6A), indicating 
that the spring bottom temperature was lower in the 
northern WEAs. Sediment explained the final 31.4% of 
the environmental correlations for the beam trawl col¬ 
lection and aligned with the first dimension and vari¬ 
ability in the northern WEAs stations (Fig. 6A). 
Latitude and longitude combined explained most of 
the environmental variability (59.2%) for the bottom 
trawl station groups and assemblages, and aligned 
with the first dimension (Fig. 6B). The importance of 
the remaining 3 environmental variables (depth, light, 
and bottom water temperature) ranged from 14.4% to 
12.9%. Bottom water temperature and depth opposed 
each other (Fig. 6B), with temperature decreasing with 
increasing depth. Light aligned near the middle of the 
2 dimensions (Fig. 6B), and was related to day-night 
variability in bottom trawl collections. Daytime col¬ 
lections were spread throughout the ordination space, 
whereas nighttime collections were clustered in the 
lower left quadrant (data not shown). 
Discussion 
The differences in the identified fish and macro-inver¬ 
tebrate assemblages in the collections made with the 2 
gear types highlight the importance of employing multi¬ 
ple types of gears for environmental assessments (Wilson 
et ah, 2010). The beam trawl collected a higher propor¬ 
tion of juvenile fish, small noncommercial fish, and small 
macro-invertebrate prey species than the bottom trawl 
(Suppl. Table (online only); Figs. 2 and 3). The diversity 
of the beam trawl samples may be the reason why the 
ordinations for those stations were more variable than 
the bottom trawl stations. In contrast, the bottom trawl 
caught a higher average proportion of and larger (i.e., 
older) demersal and pelagic, commercially important fish 
and fewer macro-invertebrates per station (Suppl. Table; 
Figs. 2 and 3) and generally had lower variability in or- 
