90 
Fishery Bulletin 1 13(1) 
low-intensity trawling. The densities of sessile inverte- 
brates, which peaked at 2.3 individuals 100 m-2, were 
not significantly different between trawled and control 
plots at any point over the study period. The densities 
of mobile macro-invertebrates were higher, peaking at 
220 individuals 100 m -2 but did not differ significantly 
between trawled and control plots after low-intensity 
trawling activities. 
High-intensity trawling 
Scour marks were immediately visible on the substrate 
during visual surveys conducted 2 weeks after high- 
intensity trawling — an observation consistent with 
those of the condition of the seafloor after low-intensity 
trawling. The mean percent cover of microtopographic 
complexity features (Fig. 4A) was not significantly dif- 
ferent between the trawled and untrawled plots in Au- 
gust 2010, 1 year after low-intensity trawling and ap- 
proximately one month before the directed trawling in 
the same study plots at a higher intensity. Immediately 
after high-intensity trawling, the mean percent cover 
microtopographic features in both control and trawled 
plots declined precipitously, 27% and 25% respectively, 
but did not differ statistically (Z=0.015, P=0.496). 
However, in May 2011, at 6 months after high-in- 
tensity trawling, the mean percent cover of microto- 
pographic features had increased in both trawled and 
control plots, and a significant increase of 24% in the 
control plots versus 4% in the trawled plots (Z=2.802, 
P=0.003). At 1 year after trawling, the mean percent 
cover had declined again in both types of plots to 37% 
in control plots and 27% in trawled plots and did not 
differ statistically (Z=0.675, P=0.251). In May 2012, 
approximately 1.5 years after trawling, mean percent 
cover had declined further in both control and trawled 
plots to 22.2% and 19.4%, respectively. The overall de- 
cline in mean percent cover of microtopographic fea- 
tures after the cumulative impact of both low- and 
high-intensity trawling was -53.96% and -52.43% in 
control and trawled plots, respectively. 
The densities of both structure-forming (Fig. 4B) and 
mobile (Fig. 40 macroinvertebrate organisms continued 
to be very low in both trawled and control plots after 
high-intensity trawling. The densities of structure-form- 
ing invertebrates, which peaked at 0.026 individuals 100 
m -2 in the control plots, were not significantly differ- 
ent between trawled and control plots at any point of 
the study period. The densities of mobile macroinverte- 
brates did not differ significantly between trawled and 
control plots after high-intensity trawling activities. 
Power analyses 
The results of power analyses (Fig. 5) for our revised 
sampling design for both overall comparisons (A r =1600) 
and year-to-year comparisons (A=200) indicated that 
our ability to detect an impact from trawling on mi- 
crotopographic structure with the Z-test was substan- 
tial and, therefore, that even a small effect (0.2, as per 
Lipsey, 1990) would be clearly discernible. That high 
power declined very little with the t-tests (Fig. 5) for 
identification of trawling effects on the densities of both 
structure-forming and mobile macro-invertebrates. 
Discussion 
We aimed to quantify impacts of bottom trawling on 
the structural attributes of fish habitat in unconsoli- 
dated sandy sediments on the continental shelf off cen- 
tral California. The persistence of scour marks from 
the doors of the small-footrope bottom trawl was the 
primary impact observed in our study. Some smooth- 
ing of microtopographic structure on the seafloor was 
observed in the trawled plots compared with control 
plots (as measured by reductions in the percentage of 
the seafloor that was observed to have bioturbation). 
However, the minimal differences in microtopographic 
structure were statistically significant only during one 
of 8 sampling periods over the course of this study. 
Further, even with a high statistical power to discern 
such effects, no impacts from bottom trawling were 
observed in the densities of sessile, structure-forming 
invertebrates, declines of which are a common indica- 
tor of trawling impacts on fish habitat in other stud- 
ies worldwide (Auster and Langton, 1999; Kaiser et al., 
2002; Barnes and Thomas, 2005; Hiddink et al., 2007). 
Given the considerable variability observed in the den- 
sities of mobile macro invertebrates in the study area, 
no significant differences attributable to trawling ex- 
isted between trawled and control plots. 
Microtopographic features on the seafloor have been 
shown to provide habitat for demersal fishes of a vari- 
ety of species (Auster et ah, 1995; Auster et al., 1997; 
Malatesta and Auster, 1999; Tissot et al., 2006), creat- 
ing the potential for larger, population-scale impacts 
from bottom trawling (Lindholm et al., 2001; Rooper 
et ah, 2011). Much of the global literature on the ef- 
fects of bottom trawling has focused on hard substrates 
(NRC, 2002) and associated high-relief structural 
attributes (Freese et ah, 1999; Henry et ah, 2006; 
Stone, 2006; Hiddink et ah, 2006; Althaus et ah, 2009). 
The comparatively limited structural attributes in low- 
relief, unconsolidated sediments (such as sand waves 
and depressions or burrows) still can provide impor- 
tant refugia for fishes (Auster et ah, 1997; Gerstner, 
1998; Gerstner and Webb, 1998; Sanchez et al, 2000). 
We expected that impacts to the structural attributes 
of fish habitat, including physical (microtopographic) 
and biological (densities of structure-forming inverte- 
brate) features would be discernible by a difference 
between control and trawled plots after low-intensity 
bottom trawling and that the difference would increase 
following high-intensity trawling. These predictions 
were based on our understanding of seafloor impacts 
from other studies (Auster et ah, 1996; Engel and Kvi- 
tek, 1998; Hixon and Tissot, 2007; de Marignac et ah, 
