Vous êtes sur la page 1sur 7

aqua$$$827

Aquaculture Research, 1997, 28, 43–49

Effects of decreases in photoperiod on growth and


bimodality in Atlantic salmon Salmo salar L.

O T Skilbrei
Department of Aquaculture, Institute of Marine Research, Bergen-Nordnes, Norway

T Hansen
Matre Aquaculture Research Station, Institute of Marine Research, Matredal, Norway

S O Stefansson
Department of Fisheries and Marine Biology, University of Bergen, High Technology Centre, Bergen, Norway

Correspondence: O T Skilbrei, Department of Aquaculture, Institute of Marine Research, POB 1870, N-5024 Bergen-Nordnes, Norway

response of juvenile Atlantic salmon changes in the


Abstract
range of 12–15 hours of daily light. This mechanism
Two experiments were performed to study the is probably linked to the size of the parr and is one
relative significance of the absolute daylength and important reason for the development of bimodal
the change in photoperiod on the growth and length-frequency distributions.
development of bimodality in juvenile Atlantic
salmon Salmo salar L. In Experiment A juveniles
were reared on 24 h daily light until they were
seven months old (65–82 mm in length after size Introduction
grading). They were then divided into six groups In the culture of Atlantic salmon Salmo salar L.,
and subjected to six photoperiods (6, 9, 12, 15, 18 growth and developmental rates until smolting are
and 24 h of light). In Experiment B the decrease in influenced by photoperiod (Folmar & Dickhoff 1980;
photoperiod was made in two steps. First, the day McCormick, Saunders, Henderson & Harmon 1987;
length was reduced to 18 and 21 hours, three Hoar 1988; Komourdjian, Fenwick & Saunders
months after first feeding when the weight of the 1989). Endogenous rhythms regulate growth rate
juveniles averaged 2.5 g and one group was kept and smolting in the absence of natural photoperiod
under 24 hour daily light. Two months later, each changes, which act as synchronizers of these
of these treatment groups was subdivided to produce processes (Eriksson & Lundqvist 1982). For the fish
new groups of juveniles (65–82 mm in length) farming industry, variation in percentages of one-
under 6, 9 and 12 h of daily light. Irrespective of year smolts (Knutsson & Grav 1976; Skilbrei 1991)
whether the photoperiods were reduced in one or and in the timing of smolting (Saunders & Henderson
two steps, groups held under short-day photoperiods, 1970; Duston & Saunders 1990; Stefansson,
6–12 h, grew significantly slower (Exp. A) and Björnsson, Hansen, Haux, Taranger & Saunders
showed higher proportions of lower modal group 1991) are of practical interest. Although growth
fish (Exp. B) than groups treated with long-day rates are improved by rearing the fish under
photoperiods, greater than 12 h. There were low artificially extended daylengths (Saunders,
proportions of lower modal group fish among Henderson & Harmon 1987; Stefansson, Nævdal &
juveniles larger than 75 mm at the dates of decreases Hansen 1989), a decrease in photoperiod is
in daylength irrespective of photoperiod (Exp. B, 0– necessary to initiate the smolting process which is
16%), and high or variable proportions among fish completed under increasing photoperiod (Björnsson,
smaller than 75 mm, depending on photoperiod Thorarensen, Hirano, Ogasawara & Kristinsson
(Exp B, 32–71%). It is concluded that the growth 1989).

© 1997 Blackwell Science Ltd. 43


Effects of decreases in photoperiod O T Skilbrei et al. Aquaculture Research, 1997, 28, 43–49

The reduction in photoperiod may also induce regularly. Light intensities were 120 lux immediately
growth changes which are undesirable in a culture below the surface and 70 lux at the bottom. In early
situation. There is evidence that the development of February 1991, fish of 65–82 mm (fork length)
bimodal length-frequency distributions, the result from one tank were transferred to the Institute of
of processes which separate smolting from non- Marine Research and divided into six groups. Each
smolting fish in the wild (Nicieza, Braña & Toledo group contained 110 juveniles individually marked
1991) and in culture (Simpson & Thorpe 1976; by a combination of adipose fin clips and blue spots
Thorpe 1977; Thorpe, Morgan, Ottaway & Miles (Alcian blue) in one or two of nine ventral positions
1980), is influenced by photoperiod (Villarreal, relative to the fins. From February 18 the six groups
Thorpe & Miles 1988; Stefansson, Nævdal & Hansen were reared separately in 200 l covered rectangular
1990). Elson (1957) proposed that juveniles had to fibreglass tanks and exposed to 6, 9, 12, 15, 18 or
reach a critical threshold length during the autumn 24 h of daily light (P6, P9, P12, P15, P18 and
to smolt the following spring. There are observations P24). Light was provided by 15 W fluorescent tubes
that this mechanism may operate through a giving 2.1–3.4 W m–2 10 cm below the water
relationship between threshold length and decrease surface and 1.5–2.2 W m–2 at the bottom (LI-COR
in photoperiod (Thorpe et al. 1980; Skilbrei 1991; LI1000 Underwater Sensor). The quality of the
Saunders, Duston & Benfey, 1994). freshwater, treated with NaOH to increase pH to
Groups of Norwegian Atlantic salmon parr of about 6.5 before it was supplied to the laboratories,
70–80 mm separated into lower and upper mode was further improved by adding seawater to increase
fish after being transferred from continuous to salinity to approximately 0.5‰. Mean temperatures
natural light in December and January (Skilbrei were 6.1, 6.5 and 7.2°C during the first (4 April),
1991). At that time the natural daylength was second (11 May) and third period (14 June) of
approximately 7 h (latitude 60°159) and measurement, respectively, and between 9 and 13°C
photoperiods other than continuous and natural for the rest of the experiment.
light were not examined. Even though there may From the start of the experiment in early February,
be a relationship between a critical length of about all fish were measured (fork length to the nearest
75 mm, photoperiod and bimodality the relationship mm) and weighed (total weight to nearest 0.1 gram)
between the magnitude of decrease in photoperiod five times until early September. On September 1–2
and the absolute daylength after decrease is not problems with the NaOH-treatment equipment at
known. This study reports the findings from two the Institute resulted in the death of 35% of the fish
experiments which were conducted to investigate and the experiment was terminated.
the effects of reducing the photoperiod during
autumn on the growth rates and length-frequency
distributions. The aim was to examine the effects of
Experiment B
specific photoperiods and also the importance of the
size of the reduction in photoperiod. To faciliate The experiment was performed at Matre
comparison between these two aspects and with the Aquaculture Research Station, Matredal, Western
results from the referred study (Skilbrei 1991), the Norway. Approximately 3000 fish of one family
experimental designs were standardized so as to group were reared together under continuous light
include juveniles of a narrow length range (65–82 from first feeding in mid-April. On July 2 1990,
mm) at the time of exposure to reduced photoperiods. when the fish had a mean weight of 2.5 g, they
were separated randomly into three 1 m2 fibreglass
tanks and reared under 24, 21 or 18 h of daily light
Materials and methods (P24, P21 and P18) at 260–300 lux light intensities.
Temperature was maintained at about 12°C from
Experiment A
first feeding through to late June. The maximum
Juvenile Atlantic salmon of mixed genetic origin rearing temperature was about 15°C in early August
were obtained from the commercial hatchery Salar 1990 and decreased gradually to a minimum of 6°C
A/S, Ålvik, western Norway. The fish were first in winter. On September 12 1990, fish smaller than
fed in June 1990 and then reared under elevated 65 and larger than 82 mm were removed from the
temperatures (. 10°C) and continuous light in 3 m tanks. The remaining fish (n 5 1060) were fin-
diameter fibreglass tanks. The fish were size graded clipped according to photoperiod and fork length

44 © 1997 Blackwell Science Ltd, Aquaculture Research, 28, 43–49


Aquaculture Research, 1997, 28, 43–49 Effects of decreases in photoperiod O T Skilbrei et al.

(below and above 75 mm) by using combinations (Table 1). When the photoperiod was reduced by 6
of clipping of adipose and pelvic fins (6 possible or more hours, i.e. to 12 h daily light or lower,
alternatives). A quarter of the fish were maintained there was no correlation between the reduction in
under photoperiods of 24, 21 or 18 h daily light photoperiod and the proportions of lower modal
while the rest were mixed in three additional 1 m2 group fish among the fin-clipped small (linear
fibreglass tanks and exposed to 12, 9 or 6 h of light regression on arcsin transformed data of Table 1
daily (P24–12, P24–9, P24–6, P21–12, P21–9, where photoperiod reduction is 6 h or more, slope
P21–6, P18–12, P18–9 and P18–6). Fork lengths P 5 0.52; intercept P 5 0.93) or large fish (slope
were measured on January 9 1991. The level of P 5 1.0; intercept P 5 0.06).
mortality during the period of the experiment was
1.9% (20 fish).
In both experiments a commercial dry diet
Discussion
(Skretting) was dispensed according to temperature
and fish size (Austreng, Storebakken & Åsgård 1987) The present results show that a decrease in
in both experiments using automatic feeders which photoperiod between two constant levels strongly
operated during the last 6 h of each experimental influences the subsequent growth and development
photoperiod at 7 min intervals. During of five to seven month old juvenile Atlantic salmon.
measurements the fish were anaesthetized using Further, the effect of the decrease is dependent on
benzocaine. the size of the juveniles at the date of decrease and
on the magnitude of the reduction in photoperiod.
The separation into slowly and rapidly growing
Results parr during autumn is usually observed through
the establishment of bimodal length-frequency
Experiment A distributions. The lack of bimodality in Experiment
Photoperiod affected the mean growth rate of the A may have several explanations; the experiment
groups differently, resulting in two clusters of growth may have been terminated too early or the
curves depending on whether the groups had been bimodality of the mixed genetic material may have
exposed to ‘short’; (P6–12) or ‘long’ (P15–24) been hidden because of overlap of modal groups of
photoperiods (Fig. 1). The mean lengths of the three different mean sizes as both proportions and size of
groups within each of these two categories in early the modal groups are partly under genetic control
September were not statistically different (t-tests; (Bailey, Saunders & Buzeta 1980). The feeding
P . 0.05, Fig. 1), but the mean lengths between regime (6 h per day) used in this study may also
the two categories were significantly different (P , have influenced the growth performance of the
0.001). There was no clear bimodality in P15–24, treatment groups, but the clear differences between
but the length-frequencies were skewed to the left, the short- and the long-day photoperiods and the
partly because of a small number of fish that were similarities between groups held under 6 to 12 h of
less than 120 mm in length (Fig. 2). daily light in both experiments do not indicate that
The individual lengths in September were feeding regime was a major factor which could
positively related to initial lengths in February, explain differences in growth.
especially for the three groups held under short-day From the growth differences in Experiment A (P6–
photoperiod. Additionally, the difference in length 12 vs P15–24) it was not possible to conclude
increment between P6–12 and P15–24 decreased whether the change (ù 12 h reduction) or the
with increased initial length (Fig. 3). magnitude of the new photoperiod (ø 12 h
daylength) was the most important growth
regulating factor. The corresponding results between
the 6–12 h d–1 treatments in Experiment B,
Experiment B
irrespective of the previous photoperiod, strongly
In January 1991, bimodal length-frequency groups indicate that daylengths of magnitude 12 h or less
had been established (Fig. 4). The proportion of have a pronounced effect on growth and the
lower modal group fish (, 100 mm) was higher development of bimodality. The present results,
under 6–12 h d–1 light compared with 18–24 h which suggest that the new absolute daylength is
d–1, especially for fish initially smaller than 75 mm more important as a winter signal in juvenile

© 1997 Blackwell Science Ltd, Aquaculture Research, 28, 43–49 45


Effects of decreases in photoperiod O T Skilbrei et al. Aquaculture Research, 1997, 28, 43–49

Figure 1 Growth curves for each


treatment group in Experiment A.
95 % confidence intervals of mean
length are included from June.

Figure 2 Length-frequency
distributions (n) in Experiment A on
September 1–5.

Atlantic salmon than the difference between the of developmental rates may possibly vary among
two photoperiods, differ from the results of Bromage, salmonid species.
Randall, Davies, Thrush, Duston, Carillo & Zanuy The response to photoperiod also seems to be
(1991) on maturation and timing of spawning of related to the size of the parr. The increase in length
rainbow trout and smoltification. They concluded increment with initial length in P6–12 in Experiment
that these developmental rates are not directly A and the clearly higher proportions of upper mode
controlled by photoperiod, but rather by changing fish among the initially largest fin-clipped group
daylength. Differences in photoperiodic regulation (75–80 mm) in Experiment B agree well with the

46 © 1997 Blackwell Science Ltd, Aquaculture Research, 28, 43–49


Aquaculture Research, 1997, 28, 43–49 Effects of decreases in photoperiod O T Skilbrei et al.

Table 1 Percentages of lower modal group fish (LM) among


initially small (65–74 mm) and large fin-clipped fish
(75–82 mm) in Experiment B. The reduction in the number
of hours of light for each photoperiod in September is
also given

Treatment Photoperiod LM (%) LM (%)


group reduction Small Large

P24–24 0 37.5 0.0


P21–21 0 39.5 2.4
P18–18 0 31.6 2.9
P18–12 6 53.9 4.1
P21–12 9 47.6 4.4
Figure 3 Individual length on February 12–17 versus P18–9 9 68.5 15.7
length increment until September 1–5 in Experiment A. P24–12 12 54.8 11.4
Initial lengths of fish in each treatment group (See Fig. 1 for P21–9 12 56.8 13.3
symbols) are grouped in 4 mm length intervals. Common P18–6 12 58.9 8.9
regression lines are calculated for P6–12 (R2 5 0.77, P24–9 15 71.0 4.1
P , 0.001) and P15–24 (R2 5 0.60, P , 0.01). P21–6 15 60.5 8.7
P24–6 18 55.2 6.7

Figure 4 Length-frequency distri-


butions (%) in Experiment B on
January 9 under the 6 final
photoperiods ranging from 6 to 24 h
of daily light. The bars for each of
the 6 groups summarize to 100%,
and they are shown stacked. See text
for further explanation of the
photoperiods of subgroups during the
first part of Experiment B.

proposal of Skilbrei (1988) and the results of Skilbrei The establishment of lower modes under long
(1991) that in parr of cultured stocks of Norwegian photoperiods (18–24 h in Experiment B) show that
salmon the probability of becoming upper mode fish there are also other causes of bimodality beside
is increased if a threshold length of approximately the interactions between length and photoperiod.
75 mm is attained before photoperiod is reduced Metcalfe (1994) has shown that differences in social
during the autumn. Further support for this status between individuals develop before autumn
argument is that the dates for the reductions in and that this behaviour evidently influences whether
photoperiod were arbitrarily chosen and there was an individual enters the upper mode or not. Such
no reason to believe that the length range of relationships, in addition to genetic differences, may
65–82 mm should contain high proportions of both have contributed to the differences in proportions of
potential lower and upper mode fish at those specific modal groups under long photoperiods between
dates. Daylengths of less than 12 h do not appear Experiments A and B.
to cause a further depression in growth. Based on the two clusters of growth curves in

© 1997 Blackwell Science Ltd, Aquaculture Research, 28, 43–49 47


Effects of decreases in photoperiod O T Skilbrei et al. Aquaculture Research, 1997, 28, 43–49

Experiment A, there seems to be a transition zone Acknowledgment


or threshold level that causes the juveniles to
The authors thank Svein Aleksandersen, Rolf Sundt
distinguish between ‘summer’ and ‘winter’ days,
and the staff at Matre Aquaculture Research Station
or between long and short days, in the range
for practical assistance and care of the fish and John
between 12 and 15 h of daily light. To be able
Thorpe who gave valuable comments about the
to study the growth response of specific size groups
manuscript. The authors are also grateful to BP
the reductions in photoperiod were given as single
Salar Ålvik A/S for providing some of the fish needed
or double cues. This experimental set-up deviates
for the experiments. The study was supported by
from the gradually declining natural photoperiod. the Department of Aquaculture, Institute of Marine
However, the results may still be representative of Research (IMR), Matre Aquaculture Research
the basal growth dynamics of Atlantic salmon. Station (IMR) and the Norwegian Fisheries Research
Bimodal size distributions have been observed in Council (NFFR).
wild populations (Nicieza et al. 1991) and in
groups planted as fry in a Norwegian stream
(Heggenes & Metcalfe 1991). The relationships
between parr size, photoperiod and growth are References
probably ecological mechanisms, preparing the
Austreng E., Storebakken T. & Åsgård T. (1987) Growth
parr for smolting and ensuring that the individual
rate estimates for cultured Atlantic salmon and rainbow
has a high probability of reaching a minimum
trout. Aquaculture 60, 157–160.
smolt size the next spring. Accordingly, smolts are Bailey J.K., Saunders R.L. & Buzeta M.T. (1980) Influence
of relatively homogeneous size when leaving the of parental smolt age and sea age on growth and
stream in schools (Solomon 1982; Hansen & smolting of hatchery-reared Atlantic salmon (Salmo
Jonsson 1985). In this way, the opportunity for salar). Canadian Journal of Fisheries and Aquatic Science
smolting the next spring may partly be determined 37, 1379–1386.
by the size of the parr in mid-autumn when the Berge Å.I., Berg A., Fyhn H. J., Barnung T., Hansen T. &
natural photoperiod approaches 12 h of daily light. Stefansson S.O. (1995) Development of salinity tolerance
in underyearling smolts of Atlantic salmon (Salmo salar)
The establishment of a relationship between
reared under different photoperiods. Canadian Journal of
parr size, photoperiod and smolt age has several
Fisheries and Aquatic Science 52, 243–251.
implications for the aquaculture industry. If it is Björnsson T.H., Thorarensen H., Hirano T., Ogasawara T.
a goal to reduce the mean smolt age, by maximizing & Kristinsson J.B. (1989) Photoperiod and temperature
the proportion of the upper modal group, then affect plasma growth hormone levels, growth condition
the parr should be kept under long photoperiod factor and hypoosmoregulatory ability of juvenile
until they grow larger than the threshold length Atlantic salmon (Salmo salar) during parr-smolt
(approximately 75 mm for Norwegian salmon) to transformation. Aquaculture 82, 77–91.
maintain sustainable growth rates during autumn Bromage N., Randall C., Davies B., Thrush B., Duston J.,
Carillo M. & Zanuy S. (1991) Photoperiodism and the
and winter. The present findings may also be
control of reproduction and development in farmed fish.
applied in out-of-season smolt production. Recent
In: Research for Aquaculture: Fundamental and applied
research (including the present results) have led
aspects. European Society for Comparative Physiology
to production schemes for zero-age smolts, and and Biochemistry, 13th conference, Antibes, Juan les
many Norwegian hatcheries now produce zero- Pins, 6–10 October 1991, 48 (Abstract).
age smolts if they can meet the temperature Duston J. & Saunders R.L. (1990) The entrainment role of
requirements. By combining artificial long, short photoperiod on hypoosmoregulatory and growth related
and then increasing (or long) photoperiod (Berge, aspects of smolting in Atlantic salmon (Salmo salar).
Berg, Fyhn, Barnung, Hansen & Stefansson 1995) Canadian Journal of Zoology 68, 707–715.
the fish reach smolt stage during their first Elson P. (1957) The importance of size in the change from
parr to smolt in Atlantic salmon. Canadian Fish Culturist
autumn, and they are transferred to sea-cages at
21, 1–6
least a half year earlier than one-year-smolts
Eriksson L.-H. & Lundqvist H. (1982) Circannual rhythms
produced under natural light. For practical and photoperiod regulation of growth and smolting in
purposes it is usually recommended that the parr Baltic salmon (Salmo salar L.). Aquaculture 28, 113–121.
reach 10 cm before the photoperiod is reduced to Folmar L.C. & Dickhoff W.W. (1980) The parr smolt
increase the smolt size. transformation (smoltification) and seawater

48 © 1997 Blackwell Science Ltd, Aquaculture Research, 28, 43–49


Aquaculture Research, 1997, 28, 43–49 Effects of decreases in photoperiod O T Skilbrei et al.

transformation in salmonids. A review of selected dynamics in relation to smolting of Atlantic salmon,


litterature. Aquaculture 21, 1–37 Salmo salar L. Aquaculture and Fisheries Management 25,
Hansen L.P. & Jonsson B. (1985) Downstream migration 9–20.
of hatchery-reared smolts of Atlantic salmon (Salmo Simpson T. & Thorpe J.E. (1976) Growth bimodality in the
salar L.) in the river Imsa, Norway. Aquaculture 45, Atlantic salmon. Council Meeting International Council for
237–248. the Exploration of the Sea M:22.
Heggenes J & Metcalfe N.B. (1991) Bimodal size Skilbrei O.T. (1988) Growth pattern of pre-smolt Atlantic
distributions in wild Atlantic salmon populations and salmon (Salmo salar L.): the percentile increment method
their relationships with age at smolt migration. Journal (PIM) as a new method to estimate length-dependent
of Fish Biology 39, 905–907. growth. Aquaculture 69, 129–143.
Hoar W.S. (1988) The physiology of smolting salmonids. Skilbrei O.T. (1991) Importance of threshold length and
In: Fish Physiology Vol. XIB (ed. by W.S. Hoar & D.J. photoperiod for the development of bimodal length-
Randall), pp. 275–343. Academic Press, New York, NY. frequency distribution in Atlantic salmon (Salmo salar).
Knutsson S. & Grav T. (1976) Seawater adaptation in Canadian Journal of Fisheries and Aquatic Science 48,
Atlantic salmon (Salmo salar L.) at different experimental 2163–2172.
temperatures and photoperiods. Aquaculture 8, 169–187. Solomon D.J. (1982) Smolt migration in Atlantic salmon
Komourdjian M.P., Fenwick J.C. & Saunders R.L. (1989) (Salmo salar L.) and sea trout (Salmo trutta L.). In: Salmon
Endocrine-mediated photostimulation of growth in and Trout Migration Behaviour Symposium (ed. by E.L.
Atlantic salmon. Canadian Journal of Zoology 67, Brannon & E.O. Salo), pp. 196–203. University of
1505–1509. Washington, Seattle, WA.
Stefansson S.O., Nævdal G. & Hansen T. (1989) The
Metcalfe N.B. (1994) The role of behaviour in determining
influence of three unchanging photoperiods on growth
salmon growth and development. Aquaculture and
and parr-smolt transformation in Atlantic salmon, Salmo
Fisheries Management 25, 67–76
salar L. Journal of Fish Biology 35, 237–247.
McCormick S.D., Saunders R.L., Henderson, E.B. & Harmon,
Stefansson S.O., Nævdal G. & Hansen T. (1990) Growth
P.R. (1987) Photoperiod control of parr-smolt
of different families of Atlantic salmon (Salmo salar)
transformation in Atlantic salmon (Salmo salar): changes
under three experimental photoperiods. Aquaculture 86,
in salinity tolerance, gill Na1, K1-ATPase activity, and
271–281.
plasma thyroid hormones. Canadian Journal of Fisheries
Stefansson S.O., Björnsson B.T., Hansen T., Haux C.,
and Aquatic Science 44, 1462–1468.
Taranger G.L. & Saunders R.L. (1991) Growth, parr-
Nicieza A. G., Braña F. & Toledo M.M. (1991) Development smolt transformation and changes in growth hormone
of length-bimodality and smolting in wild stocks of of Atlantic salmon (Salmo salar) reared under different
Atlantic salmon, Salmo salar L., under different growth photoperiods. Canadian Journal of Fisheries and Aquatic
conditions. Journal of Fish Biology 38, 509–523. Science 48, 2100–2108.
Saunders R.L. & Henderson E.B. (1970) Influence of Thorpe J.E. (1977) Bimodal distribution of length of
photoperiod on smolt development and growth of juvenile Atlantic salmon (Salmo salar L.) under artificial
Atlantic salmon (Salmo salar). Journal of the Fisheries rearing conditions. Journal of Fish Biology 11, 175–184.
Research Board of Canada 25, 2387–2401. Thorpe J.E., Morgan R.I.G., Ottaway E.M. & Miles M.S.
Saunders R.L., Henderson E.B. & Harmon P.R. (1987) (1980) Time of divergence of growth groups between
Extended daylength during autumn enhances growth potential 11 and 21 smolts among sibling Atlantic
of juvenile Atlantic salmon. Proceedings Annual Meeting salmon. Journal of Fish Biology 17, 13–21.
Aquaculture Association Canada 1, 32–33. Villarreal C.A., Thorpe J.E. & Miles M.S. (1988) Influence
Saunders R.L., Duston J. & Benfey T.J. (1994) of photoperiod on growth changes in juvenile Atlantic
Environmental and biological factors affecting growth salmon (Salmo salar). Journal of Fish Biology 33, 15–30.

© 1997 Blackwell Science Ltd, Aquaculture Research, 28, 43–49 49

Vous aimerez peut-être aussi