Asch and Collie: Changes in a benthic megafaunal community due to disturbance from bottom fishing 
453 
this species as one of the taxa most responsible for 
increased megafaunal production in CA-II. Along shal- 
low, disturbed transects outside CA-II, bivalves, such 
as Astarte spp. and C. borealis, were more abundant. 
The hard shells of these two species likely confer some 
resistance to disturbance by bottom fishing. Similarly, 
the unidentified infaunal bivalve species (most likely 
the razor clam Ensis directus) whose siphons were fre- 
quently seen along transects open to fishing may be 
less vulnerable to this form of disturbance because it 
resides deep in the substrate below depths affected by 
otter trawls. 
Many of the species most commonly observed at 
deep, undisturbed sites belong to the classes Anthozoa, 
Malacostraca, and Ophiuroidea, which are the classes 
that a meta-analysis has identified as the taxonomic 
groups most adversely affected by mobile fishing gear 
(Collie et al., 2000b). At deep sites, we and Collie et 
al. (2005) found that bottom fishing results in reduced 
abundance of multiple species of shrimp, brittle stars 
(e.g., Ophiopholis aculeata ), tubicolous polychaetes 
(e.g., Potamilla neglecta), and the toad crab Hyas co- 
arctatus. Another species that helps define the dis- 
similarity between disturbed and undisturbed areas 
at deep sites is the horse mussel Modiolus modio- 
lus, which is a long-lived, thin-shelled bivalve that is 
known to be sensitive to bottom fishing (Bradshaw et 
al., 2002; Collie et al., 2005). M. modiolus may play an 
important role in the ecology of Georges Bank because 
anecdotal information from fishermen indicates that 
the fishes G. morhua and M. aeglefinus congregate 
around their beds (Leach, 1998). As was the case at 
shallow sites, hard-shelled mollusks (i.e., Astarte spp. 
and B. undatum) and scavengers, such as Pagurus 
spp., are among the species characteristic of deep, 
disturbed sites. Both Pagurus spp. and B. undatum 
readily consume organisms injured by bottom fishing 
(Ramsay et al., 1998). 
The exact form of the relationship between bottom- 
fishing disturbance and the diversity of noncolonial spe- 
cies is highly dependent upon the particular diversity 
index under consideration. At both depth strata, in- 
creased bottom fishing results in a concurrent decrease 
in noncolonial species richness. However, Simpson’s 
index tends to exhibit its highest value at intermediate 
levels of bottom fishing. This latter pattern is consistent 
with the intermediate disturbance hypothesis, which 
proposes that a moderate amount of disturbance can 
augment diversity by creating more heterogeneous habi- 
tats and reducing the likelihood that a single climax 
species will dominate an area (Connell, 1978). Because 
bottom fishing is not boosting the absolute number of 
species present, as indicated by species richness trends, 
then the high value of Simpson’s index at intermediate 
disturbance levels must be due to a commensurate in- 
crease in species evenness resulting from the reduced 
dominance of a few abundant species. A similar pat- 
tern was identified in a study of the effect of scallop 
dredging on the benthic community of the Irish Sea 
(Bradshaw et al., 2002). 
Recovery period and recovery rates after disturbance 
Time periods associated with the recovery of benthic 
organisms after disturbance caused by bottom fishing 
are quite variable. Because of adaptations to high levels 
of natural disturbance, communities living in unconsoli- 
dated sand have been predicted to recover from bottom 
fishing in as little as 100 days (Collie et al., 2000b). 
However, recovery in structurally complex habitats and 
among particularly vulnerable species (i.e., species that 
are long lived, poorly adapted to withstand frequent 
natural disturbances, or highly susceptible to capture 
or removal by mobile fishing gear) may require longer 
time periods. For example, it is estimated to take at least 
15 years for several species of sponges off the northwest 
shelf of Australia to grow to a height of 25 cm (Sains- 
bury et al., 1997). Similarly, the fig sponge ( Suberites 
ficus) became more abundant in areas protected from 
bottom fishing within a period of 4.5 years (Lindholm 
et al., 2004). Deep sea corals, such as the samples of 
Desmophyllum cristagalli caught off Western Ireland and 
calculated to be 4000-5000 years old (Hall-Spencer et 
al., 2002), are the type of marine fauna requiring the 
longest recovery time. 
Some estimated recovery rates from bottom fishing 
reported in the scientific literature may be overly opti- 
mistic because of biases in the sampling design of many 
studies. Frequently, these studies involve trawling a 
small area (<50 m width) located inside a largely un- 
disturbed site. In the ecological literature, small-scale 
disturbances surrounded by large, unaffected areas 
are often referred to as type-1 disturbances. Because of 
the small spatial scale of a type-1 disturbance and the 
low temporal frequency of such disturbances, recoloni- 
zation may occur through immigration from adjacent 
undisturbed areas or vegetative growth (Auster and 
Langton, 1999; Kaiser et al., 2002). Recovery through 
localized immigration and vegetative growth requires 
less time than would be necessary if recovery were to 
occur through larval settlement and in situ reproduc- 
tion of remaining organisms within the disturbed area. 
In the case of type-2 disturbances where the ecological 
community is perturbed across large areas interspersed 
amongst small unaffected patches, recovery usually 
proceeds by the slower process of larval settlement, a 
process where most larvae originate from either distant 
areas or from the small unaffected patches. This latter 
scenario more realistically describes the recovery pro- 
cess after disturbance from large-scale bottom-fishing 
operations on Georges Bank. 
Our study is noteworthy because little research has 
been published on the long-term processes governing 
recovery from disturbance among colonial epifauna on 
the continental shelf of the northwest Atlantic. After 
the establishment of CA-II, several colonial and non- 
colonial taxa underwent successive increases and de- 
clines in abundance at site 17, thus, providing potential 
evidence of ecological succession. During 1997, bushy 
bryozoans briefly peaked in abundance at the shallow 
undisturbed site. Bushy bryozoans were able to react 
