Asch and Collie: Changes in a benthic megafaunal community due to disturbance from bottom fishing 
439 
provided adequate time for the completion of ecological 
succession. 
We evaluated fluctuations in the abundance of colonial 
and noncolonial epifauna in benthic photographs taken 
between the years 1994 and 2000 in areas of Georges 
Bank (northwest Atlantic) that have been classified 
as either disturbed or undisturbed by bottom fishing. 
Colonial epifauna are ecologically important because 
many taxa generate three-dimensional microhabitats 
that augment the structural complexity of the benthic 
environment. These microhabitats may directly benefit 
invertebrates and demersal fishes by aggregating food 
sources and providing refuge from visual predators 
(Henkel and Pawlik, 2005). Comparisons were made 
between two depth strata in order to determine how 
depth may modulate the response of epifauna to bot- 
tom-fishing disturbance. Because noncolonial species 
of megafauna were examined in both photographs and 
previously analyzed benthic samples collected at the 
same study sites with a 1-m wide “naturalist” dredge 
(a dredge design described in Eleftheriou and Holme, 
1984), trends in noncolonial epifauna distribution, abun- 
dance, and species composition are analyzed to facilitate 
comparisons between sampling techniques. The effect of 
bottom fishing on the diversity of noncolonial species is 
also examined. 
One of our study sites was located in an area where 
bottom fishing has been prohibited since December 
1994. Because this site had previously been fished with 
mobile gear, much of its resident epifauna had been 
destroyed at the beginning of our study. This situation 
allowed us to examine patterns of ecological recovery. 
Our research represents one of the first published stud- 
ies conducted in this geographic region on the long-term 
processes (i.e., greater than two years) affecting ecologi- 
cal recovery from disturbance among colonial epifauna 
on the continental shelf. 
Materials and methods 
Description of study sites 
In December 1994, bottom fishing was prohibited in 
three large sections of Georges Bank and the southern 
New England continental shelf to decrease levels of fish- 
ing mortality on Atlantic cod ( Gadus morhua ), haddock 
( Melanogrammus aeglefinus), and yellowtail flounder 
( Limanda ferruginea). Closed (fishing) area II (CA-II), 
which is the focus of the current research, is located 150 
km offshore on the Northeast Peak of Georges Bank (Fig. 
1A). The northern section of CA-II where the seafloor is 
principally covered by gravel substrate was designated 
as a habitat area of particular concern (HAPC) in 1998. 
The gravel substrate in this area forms an important 
habitat for Atlantic sea scallops {Placopecten magellani- 
cus), provides a nursery ground for juvenile G. morhua 
and M. aeglefinus, and serves as a spawning ground 
where Atlantic herring ( Clupea harengus ) deposit their 
demersal eggs (Lough et al., 1989). Sections of the gravel 
substrate contain dense patches of sponges, hydroids, 
bushy and encrusting bryozoans, colonial ascidians, 
and the lacy tubeworm ( Filograna implexa) (Collie et 
al., 2000a; Valentine et ah, 2007). 
During the year before the establishment of CA-II, 
scientists from the University of Rhode Island (URI) 
and the United States Geological Survey (USGS) trav- 
eled to this area on two occasions to conduct sidescan 
sonar surveys, collect samples of benthic megafauna, 
and record the fauna observed on the seafloor along 
transect lines by means of video and still cameras. Ini- 
tial assessments of the benthic megafauna and colonial 
epifauna at these study sites were published in Collie 
et al. (1997, 2000a). Between 1994 and 2000, the URI 
and USGS research team returned to Georges Bank 
on a nearly annual basis to track the recovery of the 
megafaunal community in CA-II and to evaluate inter- 
annual variations in community structure at other sites 
(Collie et al., 2005). 
Because some sites on Georges Bank contain scat- 
tered boulders that can damage fishing gear, certain ar- 
eas are generally avoided by fishermen and can, there- 
fore, be classified as undisturbed by bottom fishing. To 
identify which regions were disturbed by bottom fishing, 
a 10-km 2 area at six different sites was surveyed with 
a 100-Khz high-resolution sidescan sonar in April 1994. 
Sonograms from these surveys were inspected for trawl 
and dredge marks. Areas containing such marks were 
classified as disturbed, and areas with no visible signs 
of recent trawling or dredging activity were considered 
undisturbed. Sediment maps were consulted to ensure 
that only sites dominated by gravel substrate were sam- 
pled (Valentine et al., 1993). Disturbed and undisturbed 
sites were examined at shallow depths (40-50 m) and at 
deeper depths (80-90 m) in U.S. and Canadian waters 
(Collie et al., 1997). 
The original classification of study sites as disturbed 
or undisturbed was validated with data on scallop 
dredging and bottom trawling effort provided by the 
National Marine Fisheries Service (NMFS) and the 
Canadian Department of Fisheries and Oceans (DFO) 
(Table 1). For sites located within the U.S. Economic 
Exclusive Zone, data from the NMFS satellite vessel 
monitoring program were used to compute the num- 
ber of hours that scallop fishing boats spent dredging 
each square nautical mile of Georges Bank from 1998 
through 2000 (Collie et al., 2005). Before 1998, the clas- 
sification of sites as disturbed or not disturbed on the 
U.S. portion of Georges Bank was verified with NMFS 
data on scallop dredging effort which had a 10-minute 
latitude resolution. At shallow sites, both NMFS data 
sets consistently confirmed the accuracy of the original 
disturbance classifications that had been based on sid- 
escan sonar surveys. The disturbance classifications at 
Canadian sites were authenticated with data from DFO 
logbook reports from the commercial scallop fishery that 
had a 1-minute latitude resolution. Because the number 
of days fished at deep sites varied both inside individual 
study sites and between years (Collie et al., 2005), we 
decided that the classifications of sites located at the 
