568 
Fishery Bulletin 115(4) 
then exposed to simulated sunlight for 15 min by using 
a light source and a fiber optic guide aimed at the right 
eye of a fish. The left eye was covered with a light¬ 
blocking cloth. The 15-min simulated sunlight exposure 
was chosen to correspond with the time fish are left on 
deck during commercial trawl sorting operations (Da¬ 
vis and Olla, 2001; Davis and Schreek, 2005). Control 
fish were treated in kind, except that the light source 
was not turned on. Fish were subsequently returned to 
their holding tanks and separated with a barrier to al¬ 
low both control and light exposed fish to be held under 
identical conditions. 
Sunlight was simulated by using a high-intensity 
xenon lamp (Spectral Products, Putnam, CT) and its 
spectral range was -320-700 nm, which approximates 
the visible (400-700 nm) and the UV range of sunlight 
directly overhead at sea level (Lalli and Parsons, 1997). 
Light intensity exiting the fiber optic light guide was 
-2000 pmol-m _2 -s _1 (measured over 400-700 nm of spec¬ 
tral range) and simulated sunlight (2010 pmol-m _2 -s _1 ) 
and measured at Newport, Oregon, under ideal clear 
conditions at 1200 noon PST on 5 October 2007 and 
by using a IL 1700 Research Radiometer (International 
Light Technologies, Inc., Peabody, MA) equipped with a 
photosynthetically active radiation-filtered waterproof 
sensor. 
Evaluation of visual function with the use of an ERG 
To evaluate visual function by using ERG, fish were 
moved into a dark room in a light-proof container. In¬ 
dividuals were then lightly anesthetized with a buff¬ 
ered MS-222 solution (~5-mg/L) and the neuromuscular 
blocking drug gallamine triethiodide (Flaxedil, Sigma 
Chemical Co., St. Louis, MO, dose -20 mg/kg) injected 
into the caudal vein to reduce movement. Fish were 
then placed on a sling and enclosed in a light-blocking 
container placed in an acrylic box. The body of the fish 
was submerged in a manner such that only a small 
portion of the head and the eye would remain above 
water to receive the light stimulus. The container was 
supplied with flow-through seawater (12°C) and the 
gills of the fish remained aerated by means of a small 
submersible pump for water circulation. Fish were 
adapted to darkness for a minimum of 1 h before physi¬ 
ological measurements were taken. 
Teflon-coated silver wire electrodes with a silver chlo¬ 
ride electroplated coating, were used to record the ERG 
responses The recording electrode was placed lightly 
on the corneal surface and the reference electrode was 
placed on the skin over the head of the fish. The re¬ 
cording chamber was illuminated with a dim red light 
(peak wavelength 660 nm) produced by light-emitting 
diodes (LEDs); these remained on while the electrodes 
were positioned. The recording system was grounded 
by using a stainless-steel plate within the experimental 
apparatus. ERG signals were amplified (10,0QQx gain) 
with 1-Hz high pass and 1-kHz low-pass filter settings 
on a DAM50 amplifier (World Precision Instruments, 
Inc., Sarasota, FL). The signal was also filtered with 
a HumBug active electronic noise eliminator (Quest 
Scientific Instruments, Inc., North Vancouver, Canada) 
that removed 60-Hz noise and was digitized at a 1-kHz 
sampling frequency with a multifunction data acquisi¬ 
tion card (DAQCard-6024E, National Instruments Corp., 
Austin, TX). Light stimuli and all data were controlled 
by a custom program developed by Eric Warrant (Uni¬ 
versity of Lund, Lund, Sweden) for use in the LabVIEW 
graphical programming system for measurement and 
automation (National Instruments Corp.). 
A circular (3.8-cm diameter) light source (SL2420 
spot light, Advanced Illumination, Inc., Rochester, VT) 
was used to produce a white LED light stimulus, and a 
thin diffuser and collimating lens were used to produce 
an even field of illumination (±10%). An intensity con¬ 
troller (CS410, Advanced Illumination, Inc.) was used 
to control light output. The intensity controller was 
connected and controlled by the analog output of the 
data acquisition card. To extend the range of available 
light levels, a series of neutral density filters (Kodak 
Optical Products,.Eastman Kodak Co., Rochester, NY) 
were used to dim the light stimulus. 
As in previous studies (e.g., Brill et ah, 2008), we 
examined changes in retinal sensitivity to light result¬ 
ing from exposure to simulated sunlight by recording 
the summed potential of electrical signal in response 
(in volts [V]) to a range of light intensities (I) and sub¬ 
sequently used the data to construct voltage in relation 
to log light intensity response curves (V-log I). Light 
intensities were increased by 0.2 log-unit steps from 
a level with no measurable response, to a level that 
produced a max response. A light stimulus consisted 
of a train of five 200-ms light flashes delivered 200 ms 
apart. This stimulus was presented every 5 s and re¬ 
peated 5 times at each light intensity. The ERG re¬ 
sponses to the final flash of each train were recorded 
and averaged. At the conclusion of an experiment, fish 
were euthanized with either a massive overdose (>300 
mg/kg) of sodium pentobarbital (Beuthanasia-D, Merck 
Animal Health, Madison, NJ) injected into the caudal 
vein, or by immersion in a bath of clove oil where the 
clove oil solution was circulated over the gills by a 
small submersible pump. 
Initially, we compared ERG data for the left and 
right eyes of control fish (n= 4) that had not been ex¬ 
posed to simulated sunlight. Preliminary analysis indi¬ 
cated that right eyes produced a consistently stronger 
voltage signal than left eyes. Our original intention had 
been to use unexposed left eyes as ‘withm-fish’ controls 
for the exposed right eyes in the exposure recovery ex¬ 
periment. However, because of the difference in signal 
strength between left and right eyes, we abandoned 
this strategy and relied instead upon a comparison of 
right eyes between control fish and sunlight exposed 
fish after various periods of recovery. Fish exposed to 
simulated sunlight were divided in groups with recov¬ 
ery times of 2, 4, 6, and 10 weeks. Each group con¬ 
sisted of 8-10 individuals. 
In addition to voltage response data we also calcu¬ 
lated voltage percent maximum (p-max) data; for each 
