Lindholm et al. : Ecological effects of bottom trawling on fish habitat along the central California outer continental shelf 
85 
120'50'0*W 
Figure 1 
Map of the study area in waters off Point Buchon and Morro Bay in central California, where 4 con- 
trol plots (black) and 4 trawled plots (white) were sampled between 2009 and 2012 and grouped into 4 
analysis units (dotted lines). North and south reference sites (shown) were sampled in May 2012. Also 
included are the state waters boundary (black line) and the extent of the Rockfish Conservation Area 
(hatched area). 
a Trackpoint 3 acoustic positioning system (EdgeTech, 
West Wareham, MA) and the resulting coordinates 
were logged into Hypack navigational software, vers. 
2010 (Hypack, Inc., Middletown, CT). The ROV was 
“flown” over the seafloor at a mean altitude of 0.8 m 
and a speed of approximately 0.6 knots. 
A before-after-control-impact (BACI) sampling de- 
sign was employed for these visual surveys. Surveys 
of the 8 study plots were conducted with the ROV in 
September 2009, immediately before the low-intensity 
trawling treatment. Subsequent ROV surveys were 
conducted 2-weeks after trawling (November 2009), 6 
months after trawling (April 2010), and 1 year after 
trawling (September 2010). Visual surveys were again 
conducted with the ROV at 2-weeks (November 2010), 
6 months (May 2011), 1 year (September 2011), and 
1.5 years (May 2012) after the high-intensity trawling. 
During the final ROV surveys in May 2012, 3 addi- 
tional transects were surveyed at similar depths to the 
north and south of the study area (Fig. 1) to explore 
the extent to which the substrate and biological com- 
munity in the study area was reflective of the broader 
region of the central California coast. 
During each sampling period, 3 transects in each 
study plot were surveyed with an ROV. Each ROV 
transect measured approximately 300 m in length (20 
min in duration) as determined by species and habitat 
accumulation curves plotted from data collected previ- 
ously in soft sediment communities (Lindholm et al., 
2004, 2009; de Marignac et al., 2009) and from a review 
of preliminary data collected in the study area in the 
fall of 2008. Surveys were initiated on transects ran- 
domly from a set of pre-existing starting points located 
at either the northwest or southeast ends of each study 
plot. Determination of the starting point of a plot de- 
pended on local conditions at the time of the dive so 
that any effect of currents and wind was minimized. 
After the starting point of a plot was determined, a 
random number generator selected a starting point 
from among twenty potential points, and the ROV sub- 
sequently followed the 170-m isobath across the long 
axis of the study plot. 
