Vasslides and Able: Importance of shoreface sand ridges as habitat for fishes 
105 
the transect. Temperature has played an important role 
in regulating fish distribution in temperate waters (e.g., 
Colvocoresses and Musick, 1984; Able et al., 2006), and 
it may explain the variation in the species assemblages 
between mid- and late summer for both the beam and 
otter trawl. Furthermore, the temperature differences 
along a transect can also shape the species assemblages 
within a season, as seen in the significance levels for 
temperature between seasons in the otter trawl CCA 
(Table 6). In midsummer, when the CCA identified tem- 
perature as a statistically important environmental 
variable, there was a large temperature gradient from 
BHR-1 to BHR-5 and a slightly smaller change from 
BHR-7 to BHR-9. These temperature gradients are 
coincident with the three assemblages identified for the 
otter trawl midsummer samples in the CCA (Fig. 6A). 
In contrast, the late summer temperatures were fairly 
constant across the transect, and this constancy was 
reflected in the nonsignificant P-value for temperature 
in the CCA. 
Although we examined seasonal and annual temporal 
scales, episodic events can also have a dramatic effect 
on species assemblages. The study area is well known 
as a region of up welling during the summer months; it 
often has up to five upwelling events each year, typi- 
cally lasting a week or more (Glenn et ah, 2004). These 
upwelling events bring cooler water generally found 
at the offshore sampling locations onto the near-ridge 
stations. When sampling occurred during an upwell- 
ing event, similar species were captured at all stations 
along the transect (excluding BHR-1). However, in years 
where the bottom temperatures during sampling were 
higher than the average study temperature (1997, 1998, 
2001, and 2005), the near-ridge species assemblage 
included weakfish and northern puffer, species more 
commonly associated with late summer when tempera- 
tures are warmer (Fig. 6C). It is interesting to note that 
upwelling events of 2001 were some of the most intense 
recorded during a 9-year study (Glenn et ah, 2004). 
However, when the samples of 2001 were collected, the 
water temperatures had returned to the seasonal norm, 
thus illustrating the rapidity with which upwelling 
events break down and fish assemblages can change. 
Although dissolved oxygen appears to be a signifi- 
cant factor in the arrangement of species assemblages 
along the sampling transect, its importance may be 
confounded by its relationship with temperature and 
depth. As expected, the highest mean dissolved oxygen 
levels were found at the stations with the coldest mean 
temperatures, which were also the deepest stations. 
However, the lowest mean dissolved oxygen value was 
found at a station in the same assemblage that had the 
highest value; thus the importance of dissolved oxygen 
remains unclear. 
The trend toward less well-defined species assem- 
blages when the environmental gradients were less 
pronounced lends some support to the idea that cross 
shelf patterns in species distributions are attributable 
to environmental gradients (Jaureguizar et al., 2006; 
Walsh et al., 2006). However, the importance of habitat 
complexity in the analyses of assemblages from both 
gear types (Table 6) indicates selection of specific habi- 
tats by some species within a large-scale environmental 
gradient (see Stoner and Abookire, 2002). This has been 
shown in many laboratory and field experiments for flat- 
fishes (Neuman and Able, 1998; Stoner and Abookire, 
2002) and other demersal fishes (Sullivan et al., 2000; 
Diaz et ah, 2003). As suggested by Mueter and Norcross 
(1999), this difference may be due to differences in how 
juvenile or small fishes use benthic habitat compared to 
larger adult fishes. 
The selection of habitat within the study area changed 
with ontogenetic stage; this is particularly true for the 
sandy substrate found on the top of the ridge. The beam 
trawl samples, which contained smaller, presumably 
younger juveniles, had greater species richness and 
abundance values on the ridge top than did the otter 
trawl, which captured larger juveniles and adults. The 
sandy substrate on the top of the ridge provided im- 
portant habitat for species that bury themselves, such 
as northern stargazer (Asti'oscopus guttatus) (Able and 
Fahay, 1998) and snakefish ( Trachinocephalus myops) 
(Sulak, 1990). These species were found only on the top 
of Beach Haven Ridge, although admittedly in small 
numbers. Sand lances, another group that buries itself 
in sand, was also found predominantly in the sandy 
substrates. In a paired video sled and beam trawl sur- 
vey (where a video camera sled was towed along the 
bottom and then a beam trawl was dragged along the 
same area) on sand ridges off the coast of Maryland 
and Delaware, a substantially larger number of sand 
lances were captured in the video sled than in the beam 
trawl (Diaz et al., 2003), indicating that sand lance may 
be more important to the assemblage at Beach Haven 
Ridge than expected from the trawl results. 
Time of day also affects the abundance, species rich- 
ness, and identity of species captured in various habi- 
tats. A study of sand ridges offshore of Maryland and 
Delaware found that when complex habitats were lo- 
cated in proximity to simple habitats, fish abundance 
was twice as great in the complex habitats during the 
day, and the pattern was reversed at night (Diaz et al., 
2003) . This pattern is most likely due to 1) changes in 
foraging behavior over a diel period and to 2) smaller 
demersal fish selecting refuge from predators in complex 
habitats. The fact that all of the trawls in this study 
were conducted during daytime may explain why the 
stations located on either side of the top of the ridge, 
which had more complex habitats, had the highest val- 
ues for abundance and richness. 
There are a number of other possible explanations 
for the patterns in species abundance and assemblages 
identified herein that were not explored as part of this 
study. Investigations into the abundance and distribu- 
tion of planktonic larvae around Beach Haven Ridge 
have revealed physical processes as important mecha- 
nisms in concentrating mollusk larvae on either side of 
the sand ridge (Ma et al., 2006). These same processes 
may be causing the increased abundance of pelagic 
fish species seen on the flanks of the sand ridges. The 
