You are on page 1of 15

Aquaculture 270 (2007) 390 404

www.elsevier.com/locate/aqua-online

The effect of spectral composition and light intensity on melatonin,


stress and retinal damage in post-smolt Atlantic salmon, Salmo salar
Herve Migaud , Mairi Cowan, John Taylor, Hugh W. Ferguson
Institute of Aquaculture, University of Stirling, Stirling, Scotland, FK9 4LA, UK
Received 8 December 2006; received in revised form 20 April 2007; accepted 24 April 2007

Abstract

Metal halide lights are currently used as standard in commercial Atlantic salmon sea cages as a means of enhancing productivity
through grilse inhibition. However, such systems create bright point light sources that are neither environment specific nor species
specific and could potentially compromise fish welfare. Light emitting diodes (LEDs) are a new form of lighting technology
currently being developed for the fish farming industry that can be tuned to environment and species sensitivities through narrow
bandwidth outputs. However, prior to implementing these new high energy alternatives, any potential adverse effects must be
determined in fish. The objectives of this study were thus (1) to determine the effect of increasing intensities of blue LED light
(0.1992.7 W m 2, at 0.1 m from the light source) on light perception and stress response, and (2) to examine potential retinal
damage under these conditions in post-smolt Atlantic salmon, Salmo salar. A white LED light was also tested, as well as a very
high intensity metal halide positive control. Results demonstrated firstly that salmon perceived blue LED light (basal melatonin
levels maintained) irrespective of intensity. Secondly, fish exposed to high intensity blue LED light showed an increase in plasma
cortisol and glucose levels within 3 h, returning to a basal state 24 h post-light onset. This typical acute stress response was not
observed in fish exposed to the white LED light and lower blue light intensities which could indicate differential sensitivities to
spectral content of the light. No effects on the non-specific immune system (lysozyme activity) were observed. Finally, extensive
histological examination of the retina from fish exposed to these various light treatments revealed no signs of damage. This
demonstrates the efficiency of the adaptive mechanisms to light developed in fish.
2007 Elsevier B.V. All rights reserved.

Keywords: Salmo salar; Artificial light; LED; Stress; Retina damage; Photoreceptors

1. Introduction somatic growth resulting in decreased growth and


reduced flesh quality (Hansen et al., 1992; Oppedal et
Early maturation is a significant problem in the al., 1997; Porter et al., 1999; Endal et al., 2000).
salmon farming industry as fish channel most of their Photoperiod manipulation is an efficient tool to over-
energy reserves into gonadal development instead of come this problem (Hansen et al., 1992; Taranger et al.,
1998, 1999; Endal et al., 2000; Bromage et al., 2001)
and as such constant lighting regimes are routinely
Corresponding author. Tel.: +44 01786 467886; fax: +44 01768 applied in salmon cages (Bromage et al., 2001).
472133. Melatonin is known to be the key light perception
E-mail address: hm7@stir.ac.uk (H. Migaud). hormone produced and released by the pineal gland in
0044-8486/$ - see front matter 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.aquaculture.2007.04.064
H. Migaud et al. / Aquaculture 270 (2007) 390404 391

fish as in higher vertebrates with a daily light/dark bright point source of light, involve high running costs
rhythm (Ekstrom and Meissl, 1997; Porter et al., 1999; and much of their light energy is wasted in the form of
Bromage et al., 2001). In Atlantic salmon, the minimum unsuitable wavelengths (i.e. longer wavelength yellow-
light intensity required to suppress the rhythmic red light) which are rapidly absorbed in the water
melatonin production is 0.016 W m 2 (Migaud et al., column and therefore cannot be detected by fish (Loew
2006). However, due to the characteristics of lighting and McFarland, 1990; Migaud et al., 2006). This does
systems used and the strong light attraction behaviour of however depend to a large extent on the type of water
salmon demonstrated in several studies (Juell et al., body and season. Light emitting diodes (LEDs), a new
2003; Juell and Fosseidengen, 2004; Johansson et al., form of lighting technology being developed, can be
2006; Oppedal et al., 2007), fish in a cage will be manufactured to output specific wavelengths and thus
exposed to a wide range of light intensities depending can be matched to environment and species. In
on their swimming behaviour within the cage system particular, it has been suggested that LEDs focusing
(Juell et al., 2003; Juell and Fosseidengen, 2004). on the blue-green spectrum will be more suitable as
The manipulation of environmental parameters such these wavelengths generally penetrate seawater more
as temperature or light unavoidably results in an abrupt efficiently. Blue light ( 450 nm) has higher energy
change to rearing conditions which may result in a stress content and as such is able to penetrate deeper within the
response compromising fish welfare and overall growth water, reaching depths of up to 150 m in the clearest of
performances (Barton and Iwama, 1991; Wendelaar waters (Lalli and Parsons, 1995). In addition in vitro and
Bonga, 1997). Chronic stress events may also affect the in vivo studies have suggested trout Onchorhynchus
long-term physiology of the fish, suppressing immune mykiss and sea bass Dicentrarchus labrax are more
function (Pickering and Pottinger, 1987, 1989; Picker- sensitive to wavelengths peaking at 450500 nm (Max
ing, 1993; Harris and Bird, 2000), growth (McCormick and Menaker, 1992; Bayarri et al., 2002; Migaud et al.,
et al., 1998; Van Weerd and Komen, 1998; Gregory and unpublished). Furthermore, LEDs have lower power
Wood, 1999; Weil et al., 2001) and reproduction requirements, electrical running costs and a longer life
(Schreck et al., 2001), and can ultimately determine span than standard metal halide bulbs. Narrow band-
survival. However, it is thought that fish are able to width light using such new technologies and especially
acclimate to persisting stress events with plasma cortisol high energy short wavelength could thus provide much
returning to basal levels following the initial stress more efficient lighting systems than those currently used
response (Pickering and Pottinger, 1989). in the salmon farming. However, such short wave-
Although light has been used in the salmon industry lengths of light are considered to be much more harmful
for a number of years, to our knowledge, any potential than longer wavelengths in higher vertebrates (Young,
welfare impact of the use of high energy artificial point 1988; Dawson et al., 2001). Dawson et al. (2001) found
source lights has not been studied to date. Three that at irradiances of greater than 30 J cm 2, blue LED
important areas of concern have been identified, the light caused retinal photoreceptor damage in young
stress response, impact on the immune system and rhesus macaque monkeys. This could be a major
potential eye damage, all of which could compromise concern in salmon due to the strong phototactic
fish welfare. Preliminary data have shown, in Atlantic behaviour demonstrated in this species, these fish
salmon post-smolts reared in tanks that continuous could potentially be exposed to very high irradiance
artificial lighting regimes could result in chronic levels when swimming close to the lamps.
elevations in plasma cortisol levels for up to 3 The ultrastructure of the fish retina has been well
4 weeks (Migaud et al., unpublished data). This was characterised but its sensitivity to light remains unclear.
accompanied by a trend for reduced feed intake, a result In the fish retina, light is absorbed by photopigments
in line with the commonly reported growth dip in the that are contained within two types of photoreceptor
salmon industry. Since salmon are visual feeders and cells; rods and cones. Rods are associated with vision
attracted to light, it is hypothesised that high light under low illumination and are present in large numbers
intensities could induce retinal damage thus impairing or with larger outer segments in deep sea fish (Wagner
the salmon's ability to feed normally. Further investiga- and Mattheus, 2002; Wagner et al., 1998). This is
tions are clearly required. thought to be an adaptation to dim, down-welling blue
Metal halide bulbs are the present source of under- light. In the majority of fish, rods are not thought to be
water artificial lighting used in the industry, but in many capable of colour discrimination but can distinguish
aspects they are not suitable for fish farming as they are differences in brightness/intensity (Kusmic and Gual-
neither environment nor species specific. They create a tiere, 2000). By contrast, cones are involved in visual
392 H. Migaud et al. / Aquaculture 270 (2007) 390404

acuity and discrimination of colour, with each absorbing was to determine if such technology could be imple-
specific wavelengths of light. A combination of both mented within the salmon industry.
rods and cones is present in the majority of fish
(Bowmaker, 1990). Previous studies have demonstrated 2. Materials and methods
the presence of ultraviolet (UV), blue, green and red
cones and rods (containing rhodopsin) throughout the 2.1. Fish stock and initial rearing conditions
retina of newly hatched salmonids (Cheng et al., 2006).
As the fish grows the proportion of UV cones and opsin The trial was conducted in Scotland at the Machrihan-
expression varies (Cheng et al., 2006; Dann et al., 2003; ish Marine Environmental Research Laboratory (MERL,
Flamarique, 2002; Hawryshyn et al., 2003) however 55:44N, 5:44W) between the 11th April and 26th May,
they still appear to exhibit a range of spectral 2006. Groups of 35, 1+ Atlantic salmon mixed-sex post-
sensitivities. It is assumed that the sensitivity of the smolts (mean wet weight SEM = 823 130 g, Lakeland
retina will be tuned to the wavelength range of the stock) supplied by Howietoun Fisheries (Stirling, UK)
environment in which the fish lives (Bowmaker, 1990; and reared in tanks under simulated natural photoperiod
Partridge and Cummings, 1998). Unlike deep sea fish, and ambient temperature regimes, with seawater transfer
salmonids experience photopic conditions and presum- in April 2005, were randomly stocked into ten white 2 m
ably utilise this wider spectral range for vision (Bow- diameter covered tanks (volume of 1m3, 0.5 m deep,
maker, 1990). approx. initial stocking density: 18 kg m3). Tanks were
There are many similarities between the fish and part of a seawater flow-through system, filtered to 60 m,
human eye however there are also a number of key at a flow rate of approximately 50 L min1. Water
differences (Koppang and Bjerks, 2006; Wagner, 1990; temperature remained at 9 1 C. Fish were fed to
Wagner et al., 1998; Kusmic and Gualtiere, 2000). satiation on commercial salmon feed (Orion 6 mm pellet,
Firstly, fish do not have eyelids and are largely unable to Skretting, Invergordon) according to the manufacturer's
change their pupil aperture to protect the retina from guidelines via clockwork belt-feeders throughout the
high intensity light. Alternative protective mechanisms ambient daylight period. However, fish were starved for
do exist however such as migration of melanin granules approximately 12 h prior to sampling.
and photoreceptor mobility (Wagner, 1990; Allen and
Hallows, 1997; Allison et al., 2006). In response to light, 2.2. Experimental conditions
melanin granules migrate in an apical direction within
processes of the retinal pigment epithelium (RPE) and Fish were initially maintained on a 2-week acclima-
enshroud the photoreceptors (Allen and Hallows, 1997). tion period where they were exposed to simulated natural
Additionally, photoreceptors are capable of sliding into photoperiod (SNP). Artificial light was provided by two
or out of the deep recesses of the RPE (Wagner, 1990). 9-W fluorescent bulbs (Osram Dulux, S G23 energy
Moreover, unlike the mammalian retina, fish retina is saver, UK) that were located on the underside of tank lids
capable of regeneration (Johns, 1982; Cameron and and adjusted to turn on and off at dawn/dusk times every
Easter, 1995; Wu et al., 2001; Cheng et al., 2006). week throughout the trial. Intensity, measured at the
Vihtelic and Hyde (2000) studied light induced rod and water surface with a single sensor light channel watts
cone cell death (apoptosis) and regeneration in the adult meter (Skye Instruments Ltd., UK), was 0.32 watts m 2
albino zebrafish, Danio rerio. They found that rod and when illuminated and 0 W m 2 in the dark phase.
cone photoreceptor loss, followed by regeneration, Following acclimation, fish were exposed to their
occurred in fish after the onset of high intensity white respective light treatments for 4 weeks according to a
light. In a further study, it was found that the light randomised duplicated design.
induced photoreceptor loss was mainly confined to In order to determine light intensities for the
central and dorsal regions of the retina (Vihtelic et al., experiment, the spectral properties and intensity of
2006). light from a 100-W blue and 100 W white LED (control)
The objectives of our study were (1) to determine the unit were measured in a 3-m diameter, black polythene-
effects of increasing intensities of blue LED light on lined tank. Measurements were taken (submerged at
light perception by the fish (melatonin), acute and 0.25 m) at the light source and at 0.5 m horizontal
chronic stress response (cortisol, glucose) and non- increments. The spectral content was recorded using a
specific immunity (lysozyme activity) and (2) to portable spectroradiometer (Model EPP 2000c, Stellar-
examine potential retinal damage in Atlantic salmon, net Inc., USA). The intensities of the light units were
under these conditions. Ultimately, the aim of this work measured using a single sensor light channel watts meter
H. Migaud et al. / Aquaculture 270 (2007) 390404 393

set to a wavelength range of 400740 nm (Skye blood from the caudal vein of the fish and both eyes
Instruments Ltd., UK) and calibrated to National were removed and fixed in Bouins fixative. Blood
Physics Laboratory (UK) standards. Detection limit samples were taken within 5 min of netting, stored on
was 0.0001 W m2. ice and then centrifuged at 2500RPM (1200g) for
In reference to the intensity data from the blue LED 15 min. Following centrifugation, resulting plasma was
unit, six tanks (three in duplicate) each with a centrally aliquoted into eppendorfs and stored at 70 C until
positioned blue LED unit (100 W), were subjected to analysis. At pre-exposure, only 3 fish were sacrificed
continuous blue light at one of three intensities. This per tank (a total of 30 fish) whereas for the remaining
was designed to mimic the intensities that fish would be sampling periods, 5 fish were sacrificed per tank (10/
exposed to at different positions in a cage. High blue treatment). At the end of the trial, a further five fish from
LED intensity tanks (BH) were set up to mimic light each tank were sacrificed 2 h after dark onset and blood
intensity perceived by fish when positioned within 1 m withdrawn for melatonin analysis. At the same time, the
from the LED light unit in a cage, medium intensity eyes were removed from the SNP treatment fish to serve
tanks (BM) mimicked a 1.5 to 2.5 m distance and low as night controls.
intensity (BL) tanks mimicked a 2.53.5 m distance Positive controls were sampled at 2, 4 and 7 days
from the LED unit treatments. Control tanks (replicated) post onset of light exposure, where four fish were
were also set up, either subjected to continuous lighting sacrificed at each time and their eyes removed and fixed
from a 100-W white LED unit set at maximum intensity in Bouins fixative.
(WH) or to simulated natural photoperiod (SNP,
provided by two 9-W fluorescent bulbs as previously 2.4. Plasma analysis
described). A piece of plastic pipe was fastened into the
SNP tanks to mimic the presence of a light unit and thus Plasma cortisol levels were determined by radio-
rule out potential behaviour variables due to the immunoassay (RIA) according to North et al. (2006).
presence of the unit structure. Intensity within the blue The tritiated label was supplied by Amersham Biotech
and white light treatment tanks was measured at 0.1 m (UK) and a sheep anti-cortisol antibody from Diag-
intervals, while there were no fish in the tanks, to nostic Scotland (UK). Intra- and inter-coefficients of
prevent shading. variation were 2 and 11% respectively (n = 4), with a
A high light intensity positive control treatment was minimum sensitivity of 12 pg ml1. Glucose concen-
set up in a 0.15-m3 tank located at the University of tration was analysed colourimetrically using Infinity
Stirling Cottrell Aquarium in May 2006. 12 Atlantic Glucose Oxidase diagnostic kits (Alphalabs, Hamp-
salmon post-smolts (mean weight of 160 g, originated shire UK) and lysozyme activity was determined using
from Howietoun Fisheries, Lakeland stock as for the turbidity assay adapted from Lygren et al. (1999).
previous batch of fish used, Stirling, UK) were subjected Melatonin was analysed using commercially available
to a down-welling light intensity of 520 W m2 ELISA kits (IBL-Hamburg, Germany). Analytical
measured at the water surface (288 W m2 at the tank sensitivity of the kits was 3.0 pg ml1 with the intra-
bottom) from four tungsten halogen bulls (Armley and inter-assay variations of 3.8% and 10.7% respec-
500 W floodlights, UK). The tanks were part of a tively (n = 4).
seawater recirculation system with water temperature
maintained at 12 1 C. 2.5. Eye histology

2.3. Sampling regime Once the eye was removed, a small incision was
made in the sclera 90 to the right of the choroid fissure
Six sampling time points for stress response and to allow fixative penetration. Eyes were fixed overnight
retinal damage were carried out for each of the ten tanks in Bouin's fixative (less than 24 h) and then washed and
during daytime (11am2pm). These were taken 1 week transferred twice into fresh 70% ethanol where they
pre-exposure (baseline) to light treatment (following remained until processing. Eyes selected for processing
1 week of acclimation), 3 h, 24 h, 1 week, 2 weeks and included all positive control fish, all night-time sampled
4 weeks post light treatment onset on the 26th April fish and those from the experimental treatments of SNP,
2006. The sampling procedure consisted of netting the BH and WH at 24 h, 1 week, 2 weeks and 4 weeks post-
fish from the tanks into a lethal concentration of 2- exposure. Eyes were oriented using the location of the
phenoxyethanol solution (1 ml/L; Sigma). Immediately ventral choroid fissure and trimmed in a dorsalventral
after death, a heparinised syringe was used to withdraw plane to include the optic nerve. Subsequent processing
394 H. Migaud et al. / Aquaculture 270 (2007) 390404

to paraffin wax was routine and sections were stained (Fig. 1b). For each of the three named variable factors,
with haematoxylin and eosin (H&E). Possible changes ten measurements were made at the different locations
in retinal morphology were quantified by measuring the using image analysis software (Image Pro Plus, v. 4.5,
length of three variable factors; thickness of the whole Media Cybernetics, Inc. USA).
retina, thickness of the photoreceptor layer (i.e. distance
from the retinal pigment epithelium [RPE] to the nearest 2.6. Statistical analysis
membrane of the outer nuclear layer), and migration
distance of melanin granules within the RPE (Fig. 1a). Data were analysed using Minitab v.14.1 statistical
Measurements were made from a single retina from each software. Initial data were tested for normality using the
fish at a dorsal, ventral and central location (section on KolmogorovSmirnov test and for homogeneity of
the dorsal side of the optic nerve exit point) in the retina variances by Bartlett's test, and if necessary it was log-

Fig. 1. Histological sections of a retina (a) and an eye (b) sampled at day from a post-smolt Atlantic salmon under simulated natural photoperiod
(SNP). The slides illustrate retinal layers, measurements taken: RT, retina layer thickness; PR, photoreceptor layer thickness (outer nuclear layer not
included); M, height of melanin granules and the three regions (dorsal, central and ventral) on the retina were histology sections were performed.
H. Migaud et al. / Aquaculture 270 (2007) 390404 395

transformed. Data between treatments and sampling


times were compared by analysis of variance (ANOVA)
manipulated using a General Linear Model. Data from
replicates were pooled, as no significant differences
between replicates were observed. Posthoc multiple
comparisons were applied using Tukey's test. A
significance of p b 0.05 was applied to all statistical
tests. All results are presented as mean SEM.

3. Results

3.1. Tank light spectrum and intensities

In contrast to the narrow bandwidth blue LED unit


which emitted 89.5% of its total wavelength within
450499.5 nm the white LED unit displayed two
wavelength bands within the visible spectrum at 460 and
553 nm (Table 1, Fig. 2ab). Fig. 2c shows the spectral
content of the light emitted by the metal halide
floodlight.
Light intensity dispersion in the experimental tanks is
presented in Fig. 3. The total intensity measured 0.1 m
from the light source for Blue High (BH), Medium
(BM) and Low (BL) LED treatment tanks was 2.7 W
m2, 0.736 W m 2 (35% of high intensity) and 0.199 W
m2 (6.5% of high intensity treatment) respectively. In
the White High (WH) LED control tanks, light intensity
was slightly lower (2.1 W m 2, 0.1 m away the unit)
than that of the BH. Furthermore, light from the WH
LED did not penetrate as far as that from the BH
LED with a light intensity ratio of 7080% at
0.1 m from the unit, decreasing to approximately 50%
at 0.80.9 m.

3.2. Melatonin

Plasma melatonin levels reflected the difference


between the dark phase of the SNP control and the
light treatments (Fig. 4). A mean plasma melatonin
Fig. 2. Tank emission spectral profiles for IDEMA Aqua white (a) and
blue (b) LED units, in relation to distance from light through seawater
Table 1 and spectral profile emission from a 400-W floodlight in surface and at
the bottom of the tank (c) used in the high intensity control treatment.
Wavelength Percentage of wavelength band Note the different vertical axis scale. Wavelengths of principal peaks
band (nm) within visible spectrum are indicated on the graph.
Blue LED White LED
400449.5 7.3 8.2
450499.5 89.5 37.2
500549.5 3.2 17.7 concentration of 176.7 61.2 pg ml1 was measured in
550599.5 0.0 20.1 fish at night and below 20 pg ml1 at day. No significant
600649.5 0.0 12.0 difference was observed between fish exposed to BH,
650699.5 0.0 4.2 BM, BL or WH. Levels in all LED treatments were
700749.5 0.0 0.6
similar to day levels under SNP.
396 H. Migaud et al. / Aquaculture 270 (2007) 390404

tion (Fig. 5c). There did, however, appear to be a general


decrease over the first week and an increase over the
next 3 weeks in all treatments.

3.5. Eye histology

There were no visible pathological changes in the


eyes of fish kept under SNP, WH, BH and positive
control treatment throughout the course of the
investigation (Figs. 6 and 7). Spaces between some
of the ganglion cells in the retinal ganglion layer, and
clumped chromatin in the nuclei of lens cells were
observed but these observations were not considered
indicative of eye damage as they were recorded for
Fig. 3. Light intensity measured in 2 m diameter tanks illuminated with
eyes sampled from all treatments including SNP.
white (W) or blue (B) LED units at three different intensities (low, L;
medium, M; high, H), in relation to distance from light source. Overall, dorsal, central and ventral regions of the
Percentages in the legends indicate light intensity for each treatment retinas exhibited well-defined, organised layers includ-
relative to the blue high intensity treatment (BH) at 0.1 m from the ing the photoreceptor layer where the cell bodies of
light source. cones and rods were clear. Measurements of photo-
receptor layer thickness supported these observations
3.3. Growth data (Fig. 6ac). There were no significant differences in
mean thickness recorded between SNP, WH and BH
No significant differences in mean weight (ranging treatments in the three retinal regions, at each
from 780 to 900 g), length (from 420 to 440 mm) or sampling time. However mean thickness in the central
specific growth rate (from 0.3 to 0.9) were observed region of WH retinas did decrease significantly from
between treatments at the end of the trial. 24 h to 1 week light exposure (37.8% to 33.3%
respectively) although it was not significantly different
3.4. Cortisol, glucose and lysozyme to that of the SNP retinas (Fig. 6b). In positive control
retinas, there were no significant differences in the
Plasma cortisol concentrations in fish from the BH mean relative photoreceptor layer thickness between
treatment increased significantly 3 h post-light onset, sampling times (Fig. 7a). Conversely, retinas sampled
reaching a peak value of 102.9 25.4 ng ml 1 (Fig. 5a), from the night phase of SNP treatments showed
significantly higher than in fish from all other treatments
(1020 ng ml 1). 24 h post-exposure, plasma cortisol
concentrations in the BH fish returned to a level similar
to other treatment fish. 1 week post-exposure, BH fish
had a significantly lower mean concentration of plasma
cortisol (7.1 1.3 ng ml 1) than fish maintained under
SNP (23.0 6.4 ng ml 1). No more significant differ-
ences were observed with levels ranging between 10 and
20 ng ml 1.
Plasma glucose concentrations significantly in-
creased in the BH fish after 1 week exposure to light
(Fig. 5b); however, thereafter, no differences were
observed between treatment groups at any time point.
There were significant differences between treatment
groups prior to light application, where levels in the
SNP treatment group were significantly lower than Fig. 4. Plasma melatonin levels, sampled at night, in post-smolt
levels in the BM treatment group (1184.4 77.9 ng ml 1 Atlantic salmon exposed to blue low (BL), blue medium (BM), white
high (WH) and blue high (BH) constant light regimes in comparison to
and 1482.3 81.8 ng ml 1 respectively). a simulated natural photoperiod (SNP) (mean + SEM, n = 2, 3 fish/
Lysozyme activity did not differ significantly at any replicate). Superscripts denote significant differences between
of the sampling times over the course of the investiga- treatments.
H. Migaud et al. / Aquaculture 270 (2007) 390404 397

Fig. 5. Plasma cortisol levels (a), glucose levels (b) and lysozyme activity (c), in post-smolt Atlantic salmon kept under simulated natural photoperiod
(SNP), blue low (BL), blue medium (BM), white high (WH) and blue high (BH) light treatments. Data presented as mean SEM (n = 2, 5 fish/
replicate). Capital lettering denotes significant differences between sampling times for a given treatment and lower case lettering denotes significant
differences between treatments at a given time.

significantly increased photoreceptor layer thickness in regions (Fig. 6df). In retinas sampled from positive
the ventral retinal region compared to those sampled control fish, although the height of melanin granules in
under day (Figs. 8a, 9). The thickness of the layer both dorsal and central regions significantly decreased
decreased from a mean of 46.5% under darkness to from 2 to 4 days post light exposure, the level after
below a mean of 41.0% in the light exposed retinas. 7 days was not significantly different from that at 2 days
Furthermore, in the central retina, the thickness of the (Fig. 7b). There were no significant differences in
retina sampled under darkness was significantly melanin granule height in the ventral retinal region.
greater than that measured from fish sampled from There was, however, a significant reduction in the
WH at 1, 2 and 4 weeks. relative height of melanin granules (% photoreceptor
In terms of melanin protection, there was no layer) in SNP fish sampled at night (1924%) compared
significant difference in melanin granule migration to those sampled at day from SNP, BH and WH
height between SNP (sampled at day), WH and BH treatments (range from 57 to 63%) in all three retinal
treatments or sampling times in any of the retinal regions (Figs. 8b, 9). Although there was no measured
398 H. Migaud et al. / Aquaculture 270 (2007) 390404

Fig. 6. Thickness of the photoreceptor layer relative to retina layer (%) at dorsal (a), central (b) and ventral (c) retinal regions and height of melanin
granule migration relative to photoreceptor layer (%) at dorsal (d), central (e) and ventral (f) retinal regions. Measurements conducted in post smolt
Atlantic salmon exposed to simulated natural photoperiod (SNP), white high intensity LED (WH) and blue high intensity LED treatments at 24 h,
1 week, 2 weeks and 4 weeks post onset of light exposure. Data expressed as mean + SEM, (n = 2, 5 fish/replicate). Superscripts denote significant
differences between sampling times for a given treatment.

change in the height of melanin granules between 4. Discussion


constant light treatments, it was observed that there was
a greater packed density in positive control fish retinas With respect to commercial production, the use of
than SNP, WH and BH treatment retinas (data not artificial light and photoperiod regimes is widely
presented). accepted as tools for enhancing productivity within the
H. Migaud et al. / Aquaculture 270 (2007) 390404 399

Fig. 7. (a) Thickness of the photoreceptor layer relative to the retina layer (%) and (b) height of melanin granule migration relative to the photoreceptor
layer (%) at three different regions of the retina in post smolt Atlantic salmon (positive control) exposed to high intensity flood lights (N500 W m 2
measured at water surface) and sampled 2, 4 and 7 days post onset of light exposure. Data expressed as mean + SEM (4 fish/time point). Superscripts
denote significant differences between sampling times for a given retinal region.

salmon industry. At present, however, no lighting knowledge to help the implementation of these new
systems have been specifically tuned to aquaculture lighting regimes and investigate potential welfare
rearing systems or fish sensitivities and the use of concerns relating to the use of high energy bulbs within
artificial lighting regimes (metal halide bulbs) in cage on-growing cage systems. Only when the efficiency of
systems is costly. The transfer of novel lighting such systems and potential stress response and retina
technology such as LED units is promising as they damage are determined can such technology be
have cheaper running costs and narrow bandwidth light implemented within the salmon industry.
can be specified. It is common knowledge that the blue- Results showed a typical acute stress response
green end of the visible spectrum penetrates sea water through elevated levels of plasma cortisol and glucose
more efficiently than longer wavelengths (Bowmaker, in post-smolt Atlantic salmon (Wendelaar Bonga, 1997)
1990; Lalli and Parsons, 1995) and a number of reports following the onset of constant high intensity blue LED
suggest that fish are more sensitive to these (Max and light. Importantly, levels returned to basal values after
Menaker, 1992; Kusmic et al., 1993; Ekstrom and 24 h. Similar results have commonly been reported
Meissl, 1997; Bayarri et al., 2002; Migaud et al., following handling and confinement (Sumpter et al.,
unpublished data). Such new systems however, have 1985; Olsen et al., 2005; Pottinger et al., 1992), or
their limitations as irradiance will depend on the number environmental manipulations, including the seawater
of single LEDs used within one unit and ultimately the transfer of salmon smolts (Arnesen et al., 1998;
set-up of such units in cages will depend on their Damsgrd and Arnesen, 1998) and following abrupt
efficiency. The goal of this study was to develop changes in the rearing temperature (Mortensen and
400 H. Migaud et al. / Aquaculture 270 (2007) 390404

Fig. 8. Day and night comparison of (a) thickness of photoreceptor layer relative to the retina layer (%) and (b) height of melanin granule migration
relative to the photoreceptor layer (%) in post-smolt salmon at three retinal regions (dorsal, central, ventral). Night samples taken for SNP treatment at
the end of the trial (4 weeks post-light onset). Day samples taken from SNP, WH and BH are expressed as the mean of 24 h, 1 week, 2 weeks and
4 weeks sampling points (except for the central retinal region in (a) where a significant time effect was observed in the WH treatment between 24 h
and 1, 2 and 4 weeks). Data expressed as mean + SEM (n = 2, 5 fish/replicate). Superscripts denote significant differences between treatments for a
given retinal region.

Damsgrd, 1993; Arnesen et al., 1998). Conversely, in a white LED treatment (WH). Three hypotheses could
previous study (Migaud, unpublished data), a 3- to 4- explain this result. First, although intensity at the light
week chronic elevation in plasma cortisol levels was unit level was similar between blue and white LEDs (2
found following the onset of artificial light. The lighting 3 W m2), due to the spectral composition of the white
systems used in both experiments however were very light (N16% of total irradiance N 600 nm) and the fact
different in terms of set-up, spectral content and light that longer wavelengths are quickly absorbed within the
intensities (metal halide 400 W lighting bulbs placed water column (Loew and McFarland, 1990), light
above water on the side of 4 m diameter tanks vs. intensity experienced by the fish at the edge of the
submerged central blue LED lighting in 2 m diameter tank was lower under white light than blue light for the
tanks, respectively for previous and current trials) with same unit power output (0.14 vs. 0.27 W m 2 ,
overall higher light intensity readings in the previous respectively). Secondly, the light spectral content of
trial. the LED units may have a direct effect on the stress
Although cortisol levels were elevated 3 h following response with Atlantic salmon being more sensitive to
the exposure to high intensity blue LED light (BH), no shorter wavelengths. This contradicts results obtained
such response was observed in fish exposed to high by Volpato and Barreto (2001) who demonstrated that
H. Migaud et al. / Aquaculture 270 (2007) 390404 401

light becomes increasingly important within the aqua-


culture industry as a tool to improve fish performance, a
better understanding of the acute/long term effects of
light exposure is required.
Although high intensity blue LED lighting clearly
caused an acute stress response in post-smolt Atlantic
salmon, it did not appear to damage their eyes at a gross
morphological level. Histopathological examination of
post-smolt Atlantic salmon eyes failed to reveal any
significant pathological change following the onset of
exposure to constant high intensity blue or white LED
light, for up to 4 weeks. Adaptive retinomotor responses
possibly occurred however, indicated by the expansion
of the photoreceptor (PR) layer in the ventral region of
Fig. 9. Histological retina section taken from a SNP fish sampled at retina sampled at night (SNP) compared to those
night showing the greater packed density of the melanin granule layer. sampled at day from BH, WH and SNP treatments.
M, height of melanin granule layer. Photoreceptor mobility is a well known protective
mechanism against light (Boulton et al., 2001; Koppang
blue light can prevent the confinement-induced cortisol and Bjerks, 2006) as well as an adaptation to the
response in Nile tilapia. However, results in the present darkness. Supporting this idea is the fact that it was the
study are in accordance with Marchesan et al. (2005) ventral region of the night sampled retina that displayed
who showed that shorter blue wavelength induced a variable PR layer thickness. In the natural environ-
strong acute repulsion behaviour in seabass and sea ment, sunlight penetrates the water in downwelling
bream, Sparus aurata, following exposure. Obviously, beams of light, it is therefore considered that the ventral
as shown by Marchesan et al. (2005), levels of region of the retina receives the most light and may have
aggregation and attraction to light vary between species evolved a more efficient system of adaptation. Constant
in relation to both phylogenetic and ecological factors light treatment of albino rats was shown to cause high
and are also affected by the physical characteristics of levels of photoreceptor death in the superior retina
light, especially by its intensity. Thirdly, previous (central and dorsal) and only minimal damage in the
studies have shown that salmon are attracted to artificial inferior (ventral) retina (Gordon et al., 2002). The
lighting since they distribute themselves in relation to findings of Vihtelic et al. (2006) support this idea in
artificial light intensity (Oppedal et al., 2001; Juell et al., albino zebrafish which were over-exposed to both
2003; Juell and Fosseidengen, 2004; Johansson et al., lateral and downwelling light. Results indicated that
2006; Oppedal et al., 2007). In the current experiment, the most extensive photoreceptor damage occurred in
although not reported, behavioural changes recorded on the dorsal and central regions of the retina and not the
video were observed between salmon populations ventral region.
exposed to blue or white LED lighting units. Observa- In our study retina damage was assessed only
tions made from 1 week post light onset suggested a through histology, however, other techniques such as
tendency for increased activity and attraction to the short immunohistochemistry (IHC) and indirect fluorescence
wavelength light at high intensity as compared to the antibody technique (IFAT) have also been used for such
white. However, this is more likely to reflect chronic purposes (Vihtelic et al., 2000; Wu et al., 2001; Vihtelic
long-term (acclimation) rather than acute behavioral et al., 2006; Allison et al., 2006). Using IHC and IFAT,
response to spectral content of the light. Unfortunately, rod and cone cell damage can be identified through
such behavioural effects were not properly monitored immunodetection for rhodopsin or cone opsin (green,
and thus require further investigation. Although many red, or ultraviolet opsin). Vihtelic et al. (2000) suggested
studies have been carried out on the effects of light on that in albino zebrafish exposed to intense light,
larvae and juvenile fish (Dey and Damkaer, 1990; mislocalisation of opsin to more distal regions of the
Downing and Litvak, 2001, 2002; Brown et al., 2003), photoreceptor, indicated damage and furthermore, a
to our knowledge, there is only a limited amount of reduction of rhodopsin immunolabelling indicated cell
information on behavioural effects of varying lighting death. These techniques allow assessment of cell
conditions on adult fish especially with regards to the apoptosis and regeneration and would have been
spectral content of light (Boeuf and Le Bail, 1999). As beneficial in the current experiments to highlight
402 H. Migaud et al. / Aquaculture 270 (2007) 390404

potentially more subtle changes in response to light with the intensity of light, this was particularly evident
characteristics (wavelength, intensity). Overall, how- in positive control retina. It is hypothesised that the most
ever, it was clear that there was no severe light-induced important property of melanin is to absorb photons
damage in any retina examined, including the positive (Boulton et al., 2001). Techniques for analysing melanin
control. Thus the idea that high intensity blue LED light granule density could be incorporated into future studies
could cause a dramatic loss of vision in post-smolt to determine if more granules are synthesised as a
Atlantic salmon cannot be supported. response to damage. On a final note, in the current study,
Our results present an important question: why did although a very high intensity positive control was
constant intense blue light and positive constant high performed, only downwelling light was tested and not
intensity white light (up to 520 W m 2 at the water lateral light; this could also explain the absence of
surface) not cause significant morphological eye measurable retina damage especially given the positive
damage in post-smolt Atlantic salmon? Pathological phototactic behaviour of salmon. Nevertheless, these
changes to retinal pigment epithelium and pigment results are thought to be representative of what fish
redistribution due to short wavelength exposure has would naturally experience in a cage environment under
been easily demonstrated in a wide variety of experi- the current commercial lighting set ups.
ments concerning mammals (Young, 1988; Dawson et This study found no chronic stress response to LEDs
al., 2001; Armstrong et al., 1999). In particular, Dawson and no observed effect on the non-specific immune
et al. (2001) demonstrated eye damage in Rhesus system, nor was any physical eye damage observed. It is
monkeys where eyes were clamped open and exposed to concluded, within the limits of this study, that high
blue LED light. Secondly, albino zebrafish exposed to intensity blue LED lighting has no adverse effects on
7 days of intense halogen light, revealed visible retinal post-smolt Atlantic salmon vision. It is believed,
damage at the histological level (Vihtelic et al., 2006). In however, that the salmon did perceive the blue LED
answer to our question and in relation to these previous light because, even at low intensities, it significantly
experimental findings, it is suggested that the protective suppressed night-time plasma melatonin levels in
mechanisms of the post-smolt salmon eye which have comparison to controls. Prior to the application of
evolved for coping with rapid changes in light intensity LED systems in cages it must be acknowledged that
(Wagner, 1990) are fundamental in the prevention of such lighting systems have their limitations as irradiance
damage. Two important mechanisms in the fish retina will depend on the number of single LEDs used within
have been described to protect against light; photo- one unit and accordingly the set up of such units in cages
receptor mobility and migration of melanin granules will depend on their efficiency. Ultimately, as LED
(Koppang and Bjerks, 2006; Boulton et al., 2001), the technology expands however, prices will undoubtedly
former of which has already been discussed. Previous decrease and efficiency improve, as such, it may provide
experiments illustrating the role of melanin in protection a cheaper alternative to the high power consuming metal
arose from studies in rainbow trout (Allen et al., 1997; halide units currently used in cage on-growing systems.
Allison et al., 2006). In an experiment by Allison et al.
(2006), albino and normally-pigmented (retina contains Acknowledgements
melanin) rainbow trout previously held under covered
raceways were subjected to direct daylight. After The author would like to thank Dr. Ben North for
10 days exposure, the outer sections of rods in the conducting the cortisol assays and Dr. Andrew Davie for
central retina of albino trout had mostly disappeared, advice regarding the image analysis software. In
whereas in pigmented trout rod cell numbers were addition, the authors are grateful to the staff at
maintained. In our experiment we measured the relative Machrihanish Marine Environmental Research Labora-
height of melanin granule migration (expressed as a % tory for their assistance in fish husbandry. This project
height within the PR layer) and found that heights were was funded by IDEMA Aqua (Norway) and Marine
significantly reduced in retinas sampled from the night Harvest UK.
phase of SNP compared to those sampled at day from
WH, BH and SNP treatments. This confirms the idea References
that melanin acts as a protection against light and
suggests that, above a threshold intensity of light, Allen, D.M., Hallows, T.E., 1997. Solar pruning of retinal rods in
albino rainbow trout. Vis. Neurosci. 14, 589600.
melanin granule heights increase to a percentage Allison, W.T., Hallows, T.E., Johnson, T., Hawryshyn, C.W., Allen,
independent of the intensity. On further investigation D.M., 2006. Photic history modifies susceptibility to retinal
of the eyes it appeared that melanin density increased damage in albino trout. Vis. Neurosci. 23, 2534.
H. Migaud et al. / Aquaculture 270 (2007) 390404 403

Armstrong, D., Samuelson, D., Dawson, W., Hope, G., Ueda, T., Flamarique, I.N., 2002. Partial re-incorporation of corner cones in the
Fukada, S., Iwabuchi, S., Ueda, T., Koide, R., 1999. Retinal safety retina of the Atlantic salmon (Salmo salar). Vis. Res. 4,
of experimental light-emitting diode (LED) light in a monkey 27372745.
model. Laser Inst. Am. Proc. 4, 4756. Gordon, W.C., Casey, D.M., Lukiw, W.J., Bazan, N.G., 2002. DNA
Arnesen, A.M., Johnsen, H.K., Mortensen, A., Jobling, M., 1998. damage and repair in light-induced photoreceptor degeneration.
Acclimation of Atlantic salmon (Salmo salar L.) smolts to cold Invest. Ophthalmol. Vis. Sci. 43, 35113521.
sea water following direct transfer from fresh water. Aquaculture Gregory, T.R., Wood, C.M., 1999. The effects of chronic plasma
168, 351367. cortisol elevation on the feeding behaviour, growth, competitive
Barton, B., Iwama, G.K., 1991. Physiological changes in fish from ability, and swimming performance of juvenile rainbow trout.
stress in aquaculture with emphasis on the response and effects of Physiol. Biochem. Zool. 72, 286295.
corticosteroids. An. Rev. Fish Dis. 1, 326. Hansen, T., Stefansson, S., Taranger, G.L., 1992. Growth and sexual
Bayarri, M.J., Madrid, J.A., Sanchex-Vazquez, F.J., 2002. Influence of maturation in Atlantic salmon, Salmo salar L., reared in sea cages
light intensity, spectrum and orientation on sea bass plasma and at two different light regimes. Aquaculture Fish. Manag. 23,
ocular melatonin. J. Pineal Res. 32, 3440. 275280.
Boeuf, G., Le Bail, P.-Y., 1999. Does light have an influence on fish Harris, J., Bird, D.J., 2000. Modulation of the fish immune system by
growth? Aquaculture 177, 129152. hormones. Vet. Immunol. Immunopathol. 77, 163176.
Boulton, M., Rozanowska, M., Rozanowska, B., 2001. Retinal Hawryshyn, C.W., Martens, G., Allison, W.T., Anholt, B.R., 2003.
photodamage. J. Photochem. Photobiol. 64, 144161. Regeneration of ultraviolet-sensitive cones in the retinal cone
Bowmaker, J.K., 1990. Retinal structure of fishes. In: Douglas, R.H., mosaic of thyroxin-challenged post-juvenile rainbow trout
Djamgoz (Eds.), The Visual System of Fish. Chapman and Hall. (Oncorhynchus mykiss). J. Exp. Biol. 206, 26652673.
Bromage, N., Porter, M., Randall, C., 2001. The environmental Johansson, D., Ruohonen, K., Kiessling, A., Oppedal, F.,
regulation of maturation in farmed finfish with special reference Stiansen, J.-E., Kelly, M., Juell, J.-E., 2006. Effect of
to the role of photoperiod and melatonin. Aquaculture 197, environmental factors on swimming depth preferences of
6398. Atlantic salmon (Salmo salar L.) and temporal and spatial
Brown, J.A., Minkoff, G., Puvanendran, V., 2003. Larviculture of variations in oxygen levels in sea cages at a fjord site.
Atlantic cod (Gadus morhua): progress, protocols and problems. Aquaculture 254, 594605.
Aquaculture 227, 357372. Johns, P., 1982. Formation of photoreceptors in larval and adult
Cameron, D.A., Easter Jr., S.S., 1995. Cone photoreceptor regenera- goldfish. J. Neurosci. 2, 178198.
tion in adult fish retina: phenotypic determination and mosaic Juell, J., Fosseidengen, J.E., 2004. Use of artificial light to control
pattern formation. J. Neurosci. 15, 22552271. swimming depth and fish density of Atlantic salmon (Salmo salar)
Cheng, C.L., Flamarique, I.N., Harosi, F.I., Rickers-Haunderland, J., in production cages. Aquaculture 233, 269282.
Haunderland, N.H., 2006. Photoreceptor layer of salmonid fishes: Juell, J., Oppedal, F., Boxaspen, K., Taranger, G.L., 2003. Submerged
transformation and loss of single cones in juvenile fish. J. Comp. light increases swimming depth and reduces fish density of
Neurol. 495, 213235. Atlantic salmon Salmo salar L. in production cages. Aquac. Res.
Damsgrd, B., Arnesen, A.M., 1998. Feeding, growth and social 34, 469477.
interactions during smolting and seawater acclimation in Atlantic Koppang, E.O., Bjerks, E., 2006. The eye. In: Ferguson, H.W. (Ed.),
salmon, Salmo salar L. Aquaculture 168, 716. Systemic Pathology of Fish: A Text and Atlas of Comparative
Dann, S.G., Allison, W.T., Levin, D.B., Hawryshyn, C.W., 2003. Tissue Responses in Diseases of Teleosts, 2nd Edition. Scotian
Identification of a unique transcript down-regulated in the retina of Press, pp. 245265.
rainbow trout (Oncorhynchus mykiss) at smoltification. Comp. Kusmic, C., Gualtiere, P., 2000. Morphology and spectral sensitivities
Biochem. Physiol. Part B 136, 849860. of retinal and extraretinal photoreceptors in freshwater teleosts.
Dawson, W., Nakanishi-Ueda, T., Armstrong, D., Reitze, D., Micron 31, 183200.
Samuelson, D., Hope, M., Fukuda, S., Matsuishi, M., Ozawa, T., Kusmic, C., Barsanti, L., Passarelli, V., Gualtieri, P., 1993.
Ueda, T., Koide, R., 2001. Local fundus response to blue (LED and Photoreceptor morphology and visual pigment content in the
laser) and infrared (LED and laser) sources. Exp. Eye Res. 73, pineal organ and in the retina of juvenile and adult trout, Salmo
37147. irideus. Micron 24, 279286.
Dey, D.B., Damkaer, D.M., 1990. Effects of spectral irradiance on the Lalli, C.M., Parsons, T.R., 1995. Biological Oceanography: An
early development of chinook salmon. Prog. Fish Cult. 52, Introduction. Pergamon Press Ltd, pp. 2224.
141154. Loew, E.R., McFarland, W.N., 1990. The underwater visual environ-
Downing, G., Litvak, M.K., 2001. The effect of light and spectrum on ment. In: Douglas, R.H., Djamgoz (Eds.), The Visual System of
the incidence of first feeding by larval haddock. J. Fish Biol. 59, Fish. Chapman and Hall.
15661578. Lygren, B., Svier, H., Hjeltnes, B., Waagb, R., 1999. Examination of
Downing, G., Litvak, M.K., 2002. Effects of light intensity, spectral the immunomodulatory properties and the effect on disease
composition and photoperiod on development and hatching of resistance of the dietary bovine lactoferrin and vitamin C fed to
haddock (Melanogrammus aeglefinus) embryos. Aquaculture 213, Atlantic salmon (Salmo salar) for a short-term period. Fish Shell.
265278. Immunol. 9, 95107.
Ekstrom, P., Meissl, H., 1997. The Pineal Organ of Teleost Fishes, Marchesan, M., Spoto, M., Verginella, L., Ferrero, E.A., 2005.
vol. 7. Chapman and Hall, pp. 199284. Behavioural effects of artificial light on fish species of commercial
Endal, H.P., Taranger, G.L., Stefansson, S.O., Hansen, T., 2000. interest. Fish. Res. 73, 171185.
Effects of continuous additional light on growth and sexual Max, M., Menaker, M., 1992. Regulation of melatonin production by
maturity in Atlantic salmon, Salmo salar, reared in sea cages. light, darkness, and temperature in the trout pineal. J. Comp.
Aquaculture 191, 337349. Physiol. A 170, 479489.
404 H. Migaud et al. / Aquaculture 270 (2007) 390404

McCormick, S.D., Shrimpton, J.M., Carey, J.B., O'Dea, M.F., Sloan, (Oncorhynchus mykiss) in fresh water. Gen. Comp. Endocrinol. 88,
K.E., Moriyama, S., Bjrnsson, B.T., 1998. Repeated acute stress 454460.
reduces growth rate of Atlantic salmon parr and alters plasma Schreck, C.B., Contreras-Sanchez, W., Fitzpatrick, M.S., 2001. Effects
levels of growth hormone, insulin-like growth factor I and cortisol. of stress on fish reproduction, gamete quality, and progeny.
Aquaculture 168, 221235. Aquaculture 197, 324.
Migaud, H., Taylor, J.F., Taranger, G.L., Davie, A., Cerd-Reverter, J.M., Sumpter, J.P., Pickering, A.D., Pottinger, T.G., 1985. Stress-induced
Carrillo, M., Hansen, T., Bromage, N.R., 2006. Pineal gland elevation of plasma -MSH and endorphin in brown trout, Salmo
sensitivity to light intensity in salmon (Salmo salar) and sea bass trutta L. Gen. Comp. Endocrinol. 59, 257267.
(Dicentrarchus labrax): an in vivo and ex vivo study. J. Pineal Res. Taranger, G.L., Haux, C., Stefansson, S.O., Bjornsson, B.J., Walther,
41, 4252. B.T., Hansen, T., 1998. Abrupt changes in photoperiod affect age
Mortensen, A., Damsgrd, B., 1993. Compensatory growth and weight at maturity, timing of ovulation and plasma testosterone and
segregation following light and temperature manipulation of oestradiol-17 profiles in Atlantic salmon, Salmo salar. Aqua-
juvenile Atlantic salmon (Salmo salar L.) and Artic charr culture 162, 8598.
(Salvelinus alpinus L.). Aquaculture 114, 261272. Taranger, G.L., Haux, C., Hansen, T., Stefansson, S.O., Bjornsson, B.T.,
North, B., Turnbull, J.F., Ellis, T., Porter, M.J., Migaud, H., Bron, J., Walther, B.T., Kryvi, H., 1999. Mechanism underlying photoperiod
Bromage, N.R., 2006. The impact of stocking density on the effects on age at sexual maturity in Atlantic salmon, Salmo salar.
welfare of rainbow trout (Oncorhynchus mykiss). Aquaculture 255, Aquaculture 177, 4760.
466479. Van Weerd, J.H., Komen, J., 1998. The effects of chronic stress on
Olsen, R.E., Sundell, K., Mayhew, T.M., Mykiebust, R., Ring, E., growth in fish: a critical appraisal. Comp. Biochem. Physiol.
2005. Acute stress alters intestinal function of rainbow trout, On- 120A, 107112.
corhynchus mykiss (Walbaum). Aquaculture 250, 480495. Vihtelic, T.S., Hyde, D.R., 2000. Light-induced rod and cone cell death
Oppedal, F., Taranger, G.L., Juell, J.E., Fosseidengen, J.E., Hansen, T., and regeneration in the adult albino zebrafish (Danio rerio) retina.
1997. Light intensity affects growth and sexual maturation of J. Neurobiol. 44, 289307.
Atlantic salmon (Salmo salar) postsmolts in seacages. Aquat. Vihtelic, T.S., Soverly, J.E., Kassen, S.C., Hyde, D.R., 2006.
Living Resour. 10, 351357. Retinal regional differences in photoreceptor cell death and
Oppedal, F., Juell, J., Taranger, G.L., Hansen, T., 2001. Artificial light regeneration in light-lesioned albino zebrafish. Exp. Eye Res.
and season affects vertical distribution and swimming behaviour of 82, 558575.
post-smolt Atlantic salmon in sea cages. J. Fish Biol. 58, 15701584. Volpato, G.L., Barreto, R.E., 2001. Environmental blue light prevents
Oppedal, F., Juell, J.-E., Johansson, D., 2007. Thermo- and stress in the fish Nile tilapia. Braz. J. Med. Biol. Res. 34,
photoregulatory swimming behaviour of caged Atlantic salmon: 10411045.
implications for photoperiod management and fish welfare. Wagner, H., 1990. Retinal structure of fishes. In: Douglas, R.H.,
Aquaculture 265, 7081. Djamgoz, X. (Eds.), The Visual System of Fish. Chapman and
Partridge, J.C., Cummings, M.E., 1998. Adaptation of visual pigments Hall, pp. 107109.
to the aquatic environment. In: Archer, S.N., Djamgoz, M.B.A., Wagner, H.-J., Mattheus, U., 2002. Pineal organs in deep demersal
Loew, E.R., Partridge, J.C., Vallerga, S. (Eds.), Adaptive Mechan- fish. Cell Tissue Res. 307, 115127.
isms in the Ecology of Vision. Kluwer Academic Publishers. Wagner, H.-J., Frohlich, E., Negishi, K., Collin, S.P., 1998. The eyes of
Pickering, A.D., 1993. Growth and stress in fish production. deep-sea fish II. Functional morphology of the retina. Prog. Retin.
Aquaculture 111, 5163. Eye Res. 17, 637685.
Pickering, A.D., Pottinger, T.G., 1987. Crowding causes prolonged Weil, L.S., Barry, T.P., Malison, J.A., 2001. Fast growth in rainbow
leucopenia in salmonid fish, despite interrenal acclimation. trout is correlated with a rapid decrease in post-stress cortisol
Aquaculture 30, 701712. concentrations. Aquaculture 193, 373380.
Pickering, A.D., Pottinger, T.G., 1989. Stress responses and disease Wendelaar Bonga, S.E., 1997. The stress response in fish. Physiol.
resistance in salmonid fish: effects of chronic elevation of plasma Rev. 77, 591625.
cortisol. Fish Physiol. Biochem. 7, 253258. Wu, D.M., Schneiderman, T., Burgett, J., Gokhale, P., Barthel, L.,
Porter, M.J.R., Duncan, N.J., Mitchell, D., 1999. The use of cage Raymond, P.A., 2001. Cones regenerate from retinal stem cells
lighting to reduce plasma melatonin in Atlantic salmon (Salmo sequestered in the innernuclear layer of adult goldfish retina.
salar) and its effects on the inhibition of grilsing. Aquaculture 176, Invest. Opthalmol. Visual Sci. 42, 21152124.
237244. Young, R., 1988. Solar radiation and age-related macular degenera-
Pottinger, T.G., Prunet, P., Pickering, A.D., 1992. The effects of tion. Surv. Ophthalmol. 32, 252269.
confinement stress on circulating prolactin levels in rainbow trout

You might also like