Aquaculture 205 (2002) 183 – 200

www.elsevier.com/locate/aqua-online

The effects of f luctuating seasonal and constant

water temperatures on the photoperiodic
advancement of reproduction in female rainbow
trout, Oncorhynchus mykiss $
B. Davies *, N. Bromage
Institute of Aquaculture, University of Stirling, Stirling, FK9 4LA, UK

Received 27 April 2000; received in revised form 7 May 2001; accepted 16 May 2001

Abstract

A series of experiments were performed in order to determine the effects of varying water
temperature on the maturation of female rainbow trout subjected to photoperiodic manipulation.
Long – short photoperiod regimes, i.e. long days of LD18:6 followed by an abrupt change to short
days of LD6:18, were used to advance or delay maturation to summer months, producing
commercially desirable out-of-season eggs. At the same time, these fish were exposed to two
different seasonal water temperatures supplied by river (range 0 – 20.5
C) or borehole (range 7.0 –
10.5
C) sources.
The photoperiod regime was seen to have the primary effect on altering the timing of
maturation and appeared similar, irrespective of the prevailing water temperature. However, water
temperature had a modulating effect on time of maturation and ovulation; fish showing an ability to
delay the timing of final maturation and ovulation, when temperatures were at extremes, high or
low. Extreme temperatures, also, had a major effect on the later stages of ovary development and
subsequent egg quality. When the major time of spawning was advanced to July – August, when
river water temperatures regularly approach 20
C, a dysfunction in ovarian development occurred,
and no viable eggs were obtainable in these conditions. When the maximum water temperature
was reduced to 16
C, fish stripped in a normal fashion, but egg quality was significantly reduced.

$
Part of the results in this study were presented at the 4th International Symposium on the Reproductive
Physiology of Fish, UK (July 7 – 12, 1991), and the 5th International Symposium on the Reproductive Physiology
of Fish, USA (July 2 – 8, 1995).
*
Corresponding author. Present address: Northwest Fisheries Science Center, National Marine Fisheries
Service, 2725 Montlake Boulevard E., Seattle, WA 98112-2013, USA.
E-mail address: [email protected] (B. Davies).

0044-8486/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 0 4 4 - 8 4 8 6 ( 0 1 ) 0 0 6 6 5 - 2
184 B. Davies, N. Bromage / Aquaculture 205 (2002) 183–200

Possible alterations in farming practice in order to improve egg survival in such situations are
discussed. D 2002 Elsevier Science B.V. All rights reserved.

Keywords: Seasonal water temperature; Constant water temperature; Oncorhynchus mykiss

1. Introduction

It is clear that several environmental factors play a role in the control of reproduction
in rainbow trout and many other cold-water fish species, these include the yearly cycle of
change in daylength (photoperiod) and seasonal temperatures. There is a relatively large
body of information concerning the effects of photoperiod on the maturation of rainbow
trout, and protocols are now available for the successful advance or delay of spawning by
up to 6 months, or more generally, under conditions of constant temperature (Bromage et
al., 1993). However, the role of water temperature in determining spawning time has
largely been ignored, especially in conjunction with photoperiodic manipulation. Temper-
ature influences maturation in many fish species, and the relative importance of this
factor varies considerably between species (see Munro, 1990). A few studies have
indicated that temperatures can alter the timing of maturation and spawning of salmonids
(Henderson, 1963; Goryczko, 1972; Titarev, 1975; Morrison and Smith, 1986; Nakari et
al., 1987, 1988; Beacham and Murray, 1988; Johnstone et al., 1992; Pankhurst et al.,
1996; Taranger and Hansen, 1993; Pankhurst and Thomas, 1998; Taranger et al., 1999;
King and Pankhurst, 1999). In addition, extreme temperatures can cause egg over-
ripening with corresponding effects on embryogenesis and fry survival, and it is essential
that eggs undergo development and are ovulated, or artificially stripped at optimum
temperatures in order to maximize survival of the offspring (e.g., Billard and Breton,
1977; Billard and Gillet, 1981; Taranger and Hansen, 1993; Pankhurst et al., 1996;
Pankhurst and Thomas, 1998).
As the demands on our freshwater resources increase throughout the world, the
availability of optimum water supplies for commercial as well as wild fish production
will decrease. In addition, with the unpredictable effects of industrial effluent, as well as
global environmental changes, there is an urgent need to determine the possible effects of
abnormal temperatures on the reproduction of commercial and wild stocks. The aim of
such work is to develop guidelines and practical protocols to alleviate possible constraints
in production.
In the present study, a succession of experiments used long – short photoperiod regimes
(Billard, 1985; Duston and Bromage, 1987, 1988) to advance or delay maturation in
broodstock maintained on two different sources of water, river and borehole, providing
differing seasonal water temperatures (such water supplies are in common use for trout
production). The photoperiod and water temperature regimes were combined in order to
compare their relative effects on the maturational physiology and subsequent egg
production of female rainbow trout. These trials were used to determine the potential
use of such technology to produce out-of-season eggs on farms with variable water
supplies.
 B. Davies, N. Bromage / Aquaculture 205 (2002) 183–200 185

2. Methods

2.1. Maintenance of experimental animals

All experimental fish used were domesticated rainbow trout (Oncorhynchus mykiss)
with a winter spawning season, when under ambient photoperiod. All fish were a Severn
Springs (Larne, Northern Ireland) strain, and details of age and spawning history are
presented with each trial. Fish were fed a broodstock diet at rates according to feed
manufacturer’s tables (Trouw Aquaculture, Northwich, UK).
The following holding facilities were used:
. 5-m diameter circular tanks, which had an approximate water volume of 25 m3 and

were supplied with river water giving a seasonal temperature range of 0– 20.5
C.
Tanks under ambient photoperiod were housed outdoors and exposed to full daylight
( > 20,000 lx). Fish exposed to the artificial photoperiods were housed in identical tanks
enclosed in a light-proofed building. Light was supplied by two 1.8-m-long fluorescent
strip lights fixed centrally 3 m above the water’s surface (approximately 80 lx at the
water’s surface).
. 1.5  3.5-m raceways, which had an approximate water volume of 2.2 m3, supplied

from a borehole giving a relatively constant temperature range of 7.0 – 10.5
C. Raceways
under ambient photoperiod were housed outdoors and exposed to full daylight. Fish
exposed to artificial photoperiods were housed in identical raceways, which were light-
proofed. Light was supplied by one 60-W fluorescent bulb fixed centrally 1 m above the
water’s surface (approximately 80 lx at the water’s surface).
Rationale for photoperiod regimes: the photoperiod regimes used in the three trials were
designed, using data from Duston (1987) and Randall (1992), in order to accomplish the
specific experimental goals listed below.
Trial 1. To produce an advance in spawning time of approximately 3 months. Thus, an
advanced long day signal in February (abrupt rise to a long daylength of 18 hours)
followed by an advanced short day signal in May (abrupt drop to short daylength of 6
hours) was hypothesized to produce an advance of approximately 3 months compared to
fish remaining on the ambient photoperiod. These regimes are designated long – short
photoperiods since they consist of a long day signal followed by a short day signal. This
photoperiod regime along with an ambient photoperiod was combined with two temper-
ature regimes in order to compare the effects of photoperiod and temperature on
maturation.
Trial 2. To produce an advance in spawning time of approximately 4 months. Thus, an
advanced long day signal in December (abrupt rise to a long daylength of 18 hours),
followed by an advanced short day signal in March (abrupt drop to short daylength of 6
hours) was hypothesized to produce an advance of approximately 4 months compared to
fish remaining on the ambient photoperiod. This photoperiod regime was combined with
two temperature regimes in order to compare the effect of temperature on fish maturing
during the summer.
Trial 3. To produce a delay in spawning time of approximately 6 months while
maximizing the number and size of females undergoing their first spawning. This
photoperiod regime was designed to delay the maturation of two-year-old females by
186 B. Davies, N. Bromage / Aquaculture 205 (2002) 183–200

6 months (see Duston and Bromage, 1988). Thus, one-year-old females were given an
early long day signal in December followed by an early short day signal in March. This
inhibits the maturation of females as two-year olds. A repeat of this photoperiod regime,
at the beginning of their 3rd year of life, synchronizes and advances maturation by 6
months. This photoperiod regime was combined with two temperature regimes in order
to compare the effect of temperature on fish maturing during the summer.

2.1.1. Trial 1
Commencing on the 1st of February, four groups of two-year-old post-spawned female
rainbow trout (previously held on ambient water temperature and photoperiod) weighing
approximately 1 kg, with a natural spawning period of December –February, were sub-
jected to the following photoperiodic and water temperature regimes:

Group A: Ambient photoperiod on latitude 56
N; river water temperatures (Fig. 1).
Group B: Ambient photoperiod on latitude 56
N; borehole water temperatures (Fig. 1).
Group C: Long – short photoperiod, LD18:6 until May 10, then LD6:18; river water
temperatures (Fig. 1).
Group D: Long –short photoperiod, LD18:6 until May 10, then LD6:18; borehole water
temperatures (Fig. 1).

Groups A and C consisted of 250 females and 50 males held in circular tanks. Groups B
and D consisted of 40 female and 10 male individuals held in raceways. Close to
spawning, fish were checked weekly for ovulation. Fecundity calculations were made
from water-hardened eggs (Springate, 1985).

2.1.2. Trial 2
Commencing on the 21st of December, two groups of three-year-old female rainbow
trout weighing approximately 3 kg and close to spawning (previously spawned as two-
year olds, their first spawning, on ambient water temperature and photoperiod), were
subjected to the following photoperiodic and water temperature regimes:

Group A: Long– short photoperiod, LD18:6 until March 29, then LD6:18; river water
temperatures (Fig. 2).
Group B: Long– short photoperiod, LD18:6 until March 29, then LD6:18; borehole
water temperatures (Fig. 2).

Group A consisted of 32 females and 18 males held in a tank. This tank also held
120 one-year- and 30 three-year-old trouts not involved in this trial. Group B consisted
of 30 females and 5 males held in a raceway. In order to advance spawning of these
fish for the summer months, they were placed on the photoperiod and temperature
regimes prior to their second natural spawning period; thus, they all ovulated and were
stripped between December –March, and this stripping designated their second spawn-
ing. The fish then matured and spawned a third time, during the summer months; this
designated their third spawning. Close to spawning fish were checked weekly for
ovulation.
 B. Davies, N. Bromage / Aquaculture 205 (2002) 183–200 187

Fig. 1. The effect of water temperature and photoperiod on the spawning times of female rainbow trout in trial 1.
(i) Group A—river water plus ambient photoperiod. (ii) Group B—borehole water plus ambient photoperiod. (iii)
Group C—river water plus long – short photoperiod. (iv) Group D—borehole water plus long – short photoperiod.
River water temperatures are weekly averages, with the upper bar representing the maximum, and the lower bar
the minimum temperature recorded.

2.1.3. Trial 3
Commencing on the 21st of December, approximately 120 one-year-old virgin female
rainbow trout were subjected to the following photoperiod: LD18:6 until March 29, then
LD6:18 until December 18. These fish were all held in a tank with a river water supply.
188 B. Davies, N. Bromage / Aquaculture 205 (2002) 183–200

Fig. 2. The effect of river water (i) (Group A) and borehole water, (ii) (Group B) temperatures on the second and
third spawning times for the female rainbow trout exposed to the long – short photoperiod in trial 2. The
occurrence of initial releases of overripe eggs is also shown. The river water temperatures (upper graph) are
shown as weekly averages with the bar length representing the maximum and minimum temperature recorded
during each week.

This tank also held 32 three-year-old trout not involved in this trial. Commencing on the
18th of December, (approximately one year after the trial commenced) a portion of these
fish (now two-year olds) were randomly split into two groups (the remaining fish were
used in other experimental groupings not reported here) and subjected to the following
photoperiodic and water temperature regimes:

Group A: Long– short photoperiod, LD18:6 until April 9, then LD6:18; exposed to
river water temperatures until May 21, then combined river and borehole water
temperatures (Fig. 3).
Group B: Long – short photoperiod, LD18:6 until April 9, then LD6:18; borehole water
temperatures (Fig. 3).

Group A initially carried 40 females and 70 sex reversed males held in a tank. This
tank also held 31 two-year-old trout not involved in this trial. Group B consisted of 35
females and 10 sex reversed males held in a raceway. Close to spawning, fish were
 B. Davies, N. Bromage / Aquaculture 205 (2002) 183–200 189

Fig. 3. The effect of river plus borehole water (i) (Group A) and borehole water, (ii) (Group B) temperatures on
the spawning times of female rainbow trout exposed to the two-year long – short photoperiod in trial 3. The river
water plus borehole water temperatures are shown as weekly averages with the bar representing the maximum and
minimum temperature recorded during each week. A horizontal hatched bar represents the normal spawning
period for these fish when on river water.

checked weekly for ovulation. Stripped egg lots were collected in small buckets and
suspended in hatchery borehole water (taking care not to allow water to contact the
eggs) in order to equilibrate temperature prior to fertilization. Fecundity calculations
were made from water-hardened eggs (Springate, 1985). Individual samples of water
hardened egg lots were transported to the university hatchery and incubated in com-
partmentalized incubation trays suspended in hatchery troughs. Egg quality was assessed
up to hatching with dead eggs cleared in Stockard’s solution to determine embryo
development.

3. Statistics

When samples complied with the criteria for parametric testing, means were com-
pared using Student’s t-test with a pooled estimate of the variance (parameters analysed
using this test: fish body weight, egg size, total fecundity and relative fecundity). The
190 B. Davies, N. Bromage / Aquaculture 205 (2002) 183–200

Mann –Whitney U-test was used to compare samples, which deviated significantly from
normality or showed heterogeneous variances (parameters analysed using this test:
spawning date, %fertilization and %hatch).
Where more than two means were compared, preliminary analysis was done by one- or
two-way analysis of variance (ANOVA) for samples with normal distributions and
homogeneous variance. If means were found to be significantly different ( P < 0.05),
paired means were subsequently compared with Fisher’s Protected Least Significant
Difference multiple range test (Sokal and Rohlf, 1981) (parameters analysed using this
test: fish body weight, egg size, total fecundity and relative fecundity). All analyses were
performed using Minitab 7.2 (HP-UX version, Minitab).

4. Results

4.1. Trial 1

4.1.1. Spawning profiles

Comparison of the mean number of days to spawning showed 3 – 4 month advances for
the groups exposed to the long – short photoperiod (Groups C and D); both groups of fish
commenced spawning in August/September, compared to the groups matured under the
ambient photoperiod (Groups A and B), which began spawning in December ( P < 0.001;
Fig. 1). Comparisons between the spawning times of groups exposed to the same
photoperiod but held on different temperatures, showed that the fish exposed to the
borehole water temperatures spawned 3 –4 weeks in advance ( P < 0.001) of those fish
exposed to river water temperatures (Fig. 1).

4.1.2. Egg quality, fecundity and size

Overall egg survival was similar for all four groups and fell within the ranges reported
for the commercial trout industry (Bromage and Cumaranatunga, 1988). Unfortunately,
due to the use of commercially high capacity incubator jars, the survival of individual
egg lots could not be followed after water hardening. The total fecundities for all four
groups were similar. However, egg diameter decreased as spawning was advanced;
thus, Group D had the smallest eggs while Group A had the largest ones ( P < 0.001).
Spawning weight also increased progressively the later the spawning and, as a conse-
quence, relative fecundity was lowest for Group A and highest for Group D ( P < 0.001;
see Table 1).

4.2. Trial 2

4.2.1. Spawning profiles

4.2.1.1. Second spawning. Exposure to borehole water temperatures from December

significantly reduced the duration of the spawning period (28th December – 24th January)
compared with that seen for fish exposed to river water temperatures (21st December –
14th March) ( P < 0.001; Fig. 2).
 B. Davies, N. Bromage / Aquaculture 205 (2002) 183–200 191

Table 1
Total fecundity (eggs/fish), relative fecundity (eggs/kg), egg diameter (mm), and spawning weight (kg) for
broodstock in the four experimental groups in trial 1
Mean F S.E.M. (n).
Group Total fecundity Fish weight (kg) Relative Egg diameter (mm)
(eggs/fish) fecundity
(eggs/kg)
A—river 5421 F 276 (28) 3.099 F 0.094 1791 F 100 5.0 F 0.04
temperature
and ambient
photoperiod
B—borehole 5630 F 251 (28) 2.490 F 0.089+ + + 2319 F 130 + + 4.8 F 0.04 +
temperature
and ambient
photoperiod
C—river 5381 F 300 (35) 2.335 F 0.078*** 2398 F 168 ** 4.5 F 0.03***
temperature
and long – short
photoperiod
D—borehole 5582 F 269 (31) 2.144 F 0.060 ** 2690 F 169 4.3 F 0.03*** ,+ + +
temperature
and long – short
photoperiod
+
Indicates significant ( + P < 0.05, + + P < 0.01, + + + P < 0.001) difference compared to fish on the same
photoperiod but exposed to river water temperatures.
* Indicates significant ( ** P < 0.01, *** P < 0.001) difference compared to fish on same temperature but
exposed to ambient photoperiod.

4.2.1.2. Third spawning. Irrespective of the temperature regime, the long –short photo-
period, resulted in a 5– 6 months advance in spawning time compared to the December –
February spawning of the commercial broodstock held under ambient photoperiod and
river water temperatures. Comparison of the mean number of days to spawning showed no
significant difference between fish exposed to the river water or borehole water temper-
atures ( P >0.05; Fig. 2). However, during the routine checking of fish for ripeness in June
and July, it was found that several of the fish exposed to the river water temperatures
showed excessive softness of the abdomen but only expressed a small volume of overripe
eggs. These individuals were marked and subsequent dissection of one of these
individuals revealed intact ovaries containing eggs, which were translucent with a single
large red spot, thus, resembling overripe eggs, still held within the ovarian tissue. When
these eggs were eventually released and stripped, they continued to retain the appearance
of overripe eggs. The initial stripping times of the fish exhibiting such dysfunctional ovary
growth are shown in Fig. 2, but were not included in the calculations of the group’s
spawning time, since it was not clear when ovulation had occurred. The fish exposed to
the borehole water temperatures did not exhibit any such problems with dysfunctional
ovarian development.
All egg lots stripped from fish exposed to the river water temperatures, during July
and the majority stripped during September, did not fully water harden and showed poor
 192
B. Davies, N. Bromage / Aquaculture 205 (2002) 183–200
Table 2
Total and relative fecundity, egg diameter, spawning weight, and egg survival for broodstock in the two experimental groups in trial 3
Mean F S.E.M. (n).
Group Total fecundity Fish weight (kg) Relative fecundity Egg diameter %Fertilization %Hatch
(eggs/fish) (eggs/kg) (mm)
A 7560 F 406 (28) 3.207 F 0.135 2427 F 124 4.2 F 0.05 * 23.7 F 4.3 * 13.8 F 3.2 *
B 6869 F 413 (22) 2.843 F 0.104 2442 F 144 4.5 F 0.03 60.7 F 4.5 45.8 F 5.0
For %fertilization and %hatch, n was increased above those used for the other parameters measured in Group A (n = 33) and Group B (n = 24) since over-ripe lots showed
no fertilization and water-hardened eggs from these lots did not give meaningful measurements.
* Indicates significant difference between Group A and Group B ( P < 0.001 P < 0.01).
 B. Davies, N. Bromage / Aquaculture 205 (2002) 183–200 193

fertilization (all lots were subsequently discarded). The eggs stripped from fish exposed
to the borehole water temperatures showed good fertilization rates (data from individual
lots not available), similar to other commercial egg lots on the farm. Unfortunately, the
relatively small number of fish stripped resulted in small batches of eggs available for
on-growing, which made it commercially impractical to collect data on individual egg
lots.

4.2.2. Post-spawning mortality

Post-spawning mortality was high for both trial groups with less than 50% of the
females surviving to spawn a third time.

4.3. Trial 3

4.3.1. Spawning profiles

Comparison of the mean number of days to spawning showed that Groups A and B
had similar spawning times ( P > 0.05; Fig. 3). Both groups of fish did not spawn as
two-year olds in December – February, the normal spawning time for two-year-old
females maintained on an ambient photoperiod at this facility. Thus, the long – short
photoperiod resulted in a 6 – 7 month delay in the first spawning period, both groups
spawning between July – September during their third year of growth (year 2 of the
photoperiod regime: see Fig. 3). In Group A and Group B, several of the fish spawned
during April, 4 months ahead of the rest of the broodstock and these fish did not
mature again.

4.3.2. Egg parameters

The total fecundities, spawning weight and relative fecundities were similar for both
groups. However, egg diameter was significantly lower for Group A ( P < 0.001). The fish
exposed to borehole water temperatures (Group B) showed significantly higher fertiliza-
tion and hatch rates compared with Group A ( P < 0.001; see Table 2).

5. Discussion

5.1. The relative effects of photoperiod and temperature on the timing of spawning

The three trials presented here clearly demonstrate that photoperiod is the principle
determinant of the overall timing of maturation and spawning in the rainbow trout. Thus,
the long – short photoperiod regimes (giving 3 –3.5 months of long days) used were able
to alter the timing of spawning by 3– 7 months, the extent of the advance or delay
depended on the time in the season the photoperiod regime commenced. In contrast, the
prevailing water temperature had only a minimal effect on the timing of maturation and
spawning. Thus, in trial 1, for fish under the same photoperiod regime, exposure to
constant borehole water temperatures advanced spawning by only 3– 4 weeks compared
to fish exposed to the highly variable river water temperatures. Furthermore, in trials 2
and 3, groups of fish were exposed to markedly different water temperature regimes, but
194 B. Davies, N. Bromage / Aquaculture 205 (2002) 183–200

still showed similar spawning times. Collectively, these results suggest that photoperiod
is the major environmental factor controlling the timing of maturation and spawning in
these fish, and water temperature performs more of a modulatory role. However, this
hypothesis remains to be proven with controlled experiments using combinations of
constant temperature and photoperiod.

5.2. The effects of temperature on the timing of spawning

The results from trial 1 showed a significant effect of water temperature on the timing
of spawning. It can be seen that the fish on ambient temperature exhibited a delayed and
desynchronized spawning profile compared to the fish exposed to the constant borehole
temperatures. In addition, the extended initial spawning profile seen for the fish exposed to
the low ambient temperatures in trial 2, compared with fish moved to a higher constant
borehole temperature, further suggests a desynchronizing effect of cold temperatures on
maturation. Delays in spawning by cold winter temperatures have previously been
described for salmonids (Henderson, 1963; Goryczko, 1972; Titarev, 1975; Morrison
and Smith, 1986; Nakari et al., 1987, 1988; Johnstone et al., 1992). However, the
physiological mechanisms by which temperature affects maturation remain unclear.
Studies in trout and salmon have indicated that lower temperatures can reduce the
vitellogenic response to oestradiol application, as well as the rate of vitellogenin uptake
(Korsgaard et al., 1986; Tyler et al., 1987; Olin and von der Decken, 1989; Mackay and
Lazier, 1993). It seems, therefore, possible that the decreasing water temperatures
experienced during the later part of vitellogenesis, by the ambient control group in trial
1, delayed spawning by a combination of reduced vitellogenin production and sequestra-
tion. Earlier experiments by Davies and Bromage (1991) showed reduced serum calcium
levels (as an indicator of vitellogenin levels) in fish exposed to low ambient water
temperatures during November and December. However, in trial 2, spawning appeared to
be synchronized when fish were moved to warmer borehole temperatures during the
natural spawning period, when vitellogenesis would presumably be complete and final
maturation underway. Thus, temperature can also affect changes in the timing of these
final maturation processes.
In trial 1, the groups exposed to the long –short photoperiod spawned between
August –November, with the result that the fish on the river water experienced unusually
high water temperatures during maturation; this delayed spawning in this group com-
pared to fish exposed to the lower borehole water temperatures. Other studies in
salmonids have also reported the inhibition of maturation by high water temperatures
(rainbow trout—Billard and Breton, 1977; Breton et al., 1983; Pankhurst et al., 1996;
Pankhurst and Thomas, 1998; Chmilevsky, 1999; pink salmon—Beacham and Murray,
1988; Atlantic salmon—Johnstone et al., 1992; Taranger and Hansen, 1993; Taranger et
al., 1999; King and Pankhurst, 1999). The present results are consistent with the view
that these fish have an ability to delay spawning for a limited period until temperatures
drop to a threshold level (see Taranger et al., 1999). Pankhurst and Thomas (1998)
provided evidence indicating that this may be due to an alteration in the response to
GnRH. In addition, Kime (1979, 1980) suggested that the formation of steroid conjugates
could provide a mechanism by which high temperatures can delay spawning. However,
 B. Davies, N. Bromage / Aquaculture 205 (2002) 183–200 195

the mechanisms by which elevated temperature influence the timing of final maturation
and spawning in salmonids remains to be determined.
Although the prevailing temperature in trial 1 altered the times and profile of spawning,
in the subsequent two trials, no major effects of temperature on spawning time were
evident. Both trials utilized photoperiod regimes, which commenced in December in order
to advance fish to mid-summer, during the period of maximum river water temperatures.
In trial 2, fish undergoing the final stages of maturation, during these maximum temper-
atures, exhibited dysfunction in ovary development, which made it difficult to determine a
specific alteration in spawning time. For trial 3, any inhibitory effects on maturation
caused by the elevated water temperature appears to have been alleviated by mixing in
cooler borehole water during the last few months of maturation; thus, resulting in similar
spawning times for both groups. In addition, reducing the maximum temperature
experienced prior to ovulation to 16
C appeared to reduce the problems with ovarian
development occurring in fish exposed to the higher temperatures in trial 2. These findings
concur with those reported by Pankhurst et al. (1996), who indicated that ovarian
dysfunction in trout occurred at a constant temperature of 18 and 21
C but not 15
C
or below. Unfortunately, due to the absence of a river water temperature group (fish died,
due to a malfunction, prior to ovulation and are not reported here) for direct comparison in
trial 3, the effect of reducing the temperature to 16
C on spawning time and egg quality
remains to be confirmed.

5.3. The effects of temperature on egg quality

In trial 1, advancing fish to September – October, when river water temperatures were
falling, resulted in the successful production of eggs of a commercially acceptable
quality. Although egg sizes were smaller for fish exposed to the long – short photoperiod,
the total fecundities were similar to those of fish on ambient photoperiod; and since these
fish were of a smaller size at spawning, their relative fecundities were higher. Previous
studies have reported smaller eggs from photoperiodically advanced rainbow trout
(Nomura, 1962; Buss, 1982; Duston and Bromage, 1988), presumably resulting from
the shorter time available for vitellogenesis. A comprehensive study by Springate and
Bromage (1985) showed that egg size is not correlated with subsequent size or survival
of fry in rainbow trout and, therefore, the production of eggs from photoperiodically
advanced females can result in fry of similar quality to those from broodstock under
ambient photoperiod.
In contrast to trial 1, the eggs resulting from trials 2 and 3 showed large differences
in quality, depending on the temperature of the water supplies to the broodstock. For
trial 2, maturation and spawning was advanced to June –July, during the highest river
water temperatures (maximum 20
C), and major problems were encountered in the
stripping of the fish and in the subsequent egg quality; several individuals producing
eggs, which appeared to have overripened while still within the ovarian tissue. The
occurrence of unovulated eggs bearing a resemblance to overripe eggs has also been
reported for rainbow trout exposed to constant water temperatures of 18 and 21
C
during final maturation (Pankhurst et al., 1996). Over-ripening of salmonids’ eggs
involves the formation of a blastoderm-like structure permeated by oil globules (Nomura
196 B. Davies, N. Bromage / Aquaculture 205 (2002) 183–200

et al., 1974; Craik and Harvey, 1984), and this ‘red dot’ structure was clearly seen
within oocytes held in the ovary tissue of these fish. Studies on the over-ripening of
eggs have concentrated on the processes occurring after ovulation, while the eggs are
retained in the body cavity; (Nomura et al., 1974; Sakai et al., 1975; Billard and Breton,
1977; Escaffre et al., 1977; Escaffre and Billard, 1979; Breton et al., 1983; Springate et
al., 1984; Billard, 1985). Prior to ovulation, the re-absorption of atretic oocytes has been
well documented for early and later stages of egg development in trout (see Bromage
and Cumaranatunga, 1988). It has been suggested that this re-absorption process can
regulate fecundity and even arrest maturation (see Bromage and Cumaranatunga, 1988).
Descriptions of atretic mature stage 7 oocytes do not include aggregation of oil globules
to form the ‘red dot’ structure evident from the dysfunctional oocytes seen in the present
study, and these eggs were subsequently released and not re-absorbed. In addition, it is
known that elevated water temperatures increase the rate of over-ripening in ovulated
eggs (see Billard, 1985). It is, therefore, proposed that these dysfunctional oocytes
resulted from a high temperature-induced breakdown in oocyte integrity just prior to
ovulation, rather than an attempt to re-absorb atretic oocytes. This hypothesis remains to
be confirmed by the histological analysis of this process in order to determine at what
stage during final maturation and ovulation oocyte integrity is compromised.
In the subsequent trial 3, the water temperatures experienced by the fish from May
onwards were reduced (maximum 16
C) by the addition of borehole water to the river
water supply. Although these fish stripped normally, with most egg lots of normal
appearance and similar total and relative fecundities, the fertilization rates of the egg lots
were significantly reduced compared to eggs from fish maintained on borehole water
temperatures. If the number of days between checking the fish had been reduced, it may
have been possible to reduce the over-ripening of egg lots, since at higher temperatures
ovulated eggs held in the body cavity will become overripe more quickly (Billard, 1985).
However, from the viewpoint of the farmer, checking fish every 2 – 5 days is very time
consuming, and the small numbers of eggs obtained each time are difficult to handle
efficiently in a commercial hatchery. In addition, the supply of out-of-season sperm must
be reliable to service such lots, and if all-female eggs are being produced, farms would
need to carry large numbers of advanced sex reversed males. Thus, reducing temperatures
to below 16
C during final maturation, and spawning prevented the appearance of
dysfunctional ovaries, but egg quality was still severely reduced compared to those from
fish on borehole temperature. Further work is, therefore, required to ascertain the
maximum water temperature to which maturing and ovulating broodstock can be exposed
in order to produce good quality eggs. In addition, broodstock handling and spawning
techniques need to be assessed for broodstock held at elevated temperatures in order to
optimize egg production and survival.

5.4. Photoperiod regimes design

In trials 2 and 3, a 5-month advance in spawning was required in order to determine

if any problems would arise from the advancement of maturation and spawning to
coincide with the maximum seasonal water temperatures. As discussed in the Introduc-
tion, previous work using long – short photoperiods has indicated that the early exposure
 B. Davies, N. Bromage / Aquaculture 205 (2002) 183–200 197

to such photoperiods (during December) can advance spawning by up to 6 months.

Thus, in trial 2, three-year-old fish (maturing for the 2nd time) were exposed to long
days in December and advanced to July; however, the survival of fish in this trial was
very poor, many of the deaths occurring after the initial stripping, during the winter. In
order to avoid reductions in broodstock numbers due to postspawning mortality, trial 3
utilized virgin one-year-old fish, thereby, increasing the number of fish surviving to
spawn during the months of July and August.
In trial 3, exposure to a long –short photoperiod commencing in December (when the
fish were approaching 1 year of age) did not advance spawning of any of the fish during
the first year; and furthermore, none of these fish spawned as two-year olds, during
December – February, when other two-year-old broodstock under ambient conditions
normally spawn at this facility. Thus, the early exposure to the long –short photoperiod
had not only failed to produce an advance in spawning during the first year, but had also
delayed, or prevented, maturation as two-year olds. Photoperiod induced inhibition of
maturation has previously been described in both virgin trout and Atlantic salmon
(Duston and Bromage, 1988; Taranger, 1993; Bjornsson et al., 1994; Taranger et al.,
1998, 1999). It is the contention of these authors that a ‘gating’ mechanism for
maturation exists in these fish, i.e. a threshold ‘stage of development’ must be reached
before the animal can overtly respond to a stimulatory photoperiod. Thus, it is suggested
that the fish in trial 3 had not reached the threshold ‘stage of development’ required to
overtly respond to the initial stimulatory long – short photoperiod. It is probable that these
fish would have spawned during July – September the following year without the
exposure to a second period of long days in December; since it is known that fish,
which remained immature after early exposure to a stimulatory long – short photoperiod,
will subsequently mature approximately 1 year later than the originally advanced indi-
viduals (Duston and Bromage, 1988). The second stimulatory long –short photoperiod,
rather than a constant one, was included in this trial for the practical reason of ensuring
spawning synchrony; since extended periods of constant photoperiods have been shown
to desynchronize the spawning of rainbow trout broodstock (e.g., Bromage et al., 1984;
Duston and Bromage, 1986).
Irrespective of the mechanism by which spawning was delayed in the above trials, this
technique for delaying puberty in two-year-old fish is an effective tool for use on
broodstock farms since many farms discard the eggs produced by two-year-old broodstock
because they are too small to market (W.J. Baird, personal communication to N.
Bromage). The extra 7 months of broodstock growth resulting from the artificial delay
in maturation would allow eggs to reach a marketable size since, depending on the stock
used, egg size generally increases with the increase in body weight (Nomura, 1963;
Dubois et al., 1989; Bromage et al., 1990). Thus, the total loss of the eggs produced from
two-year-old female broodstock would be avoided.
In conclusion, for the combination of photoperiods and water temperatures tested, the
prevailing photoperiod acted as the major environmental cue in the control of seasonal
reproduction in the rainbow trout, with water temperature performing a modulatory role.
However, water temperature can have major effects on various maturational processes,
including spawning times, egg ripening, and egg quality, and must be considered when
using photoperiod to commercially control broodstock maturation.
198 B. Davies, N. Bromage / Aquaculture 205 (2002) 183–200

Acknowledgements

Thanks to Nick Younge for the supply of fish and Trouw Aqua UK and EU Fair award
to NB.

References

Beacham, T.D., Murray, C.B., 1988. Influence of photoperiod and temperature on timing of sexual maturity of
pink salmon (Oncorhynchus gorbuscha). Can. J. Zool. 66, 1729 – 1732.
Billard, R., 1985. Environmental factors in salmonid culture and the control of reproduction. In: Iwamoto, R.N.,
Sower, S. (Eds.), Salmonid Reproduction. Washington Sea Grant Program, University of Washington, Seattle,
pp. 70 – 87.
Billard, R., Breton, B., 1977. Sensibilite a la temperature des differentes etapes de la reproduction chez la truite
arc-en-ciel. Cah. Lab. Montereau 5, 5 – 24.
Billard, R., Gillet, C., 1981. Viellissement des ovules et potentialisation par la temperature des effets des micro-
pollutants du milieu aqueux sur les gametes chez la truite. Cah. Lab. Montrereau 12, 35 – 42.
Bjornsson, B.Th., Taranger, G.L., Hansen, T., Stefansson, S.O., Haux, C., 1994. The interrelation between
photoperiod, growth hormone, and sexual maturation of adult Atlantic salmon (Salmo salar). Gen. Comp.
Endocrinol. 93, 70 – 81.
Breton, B., Fostier, A., Zohar, Y., Le Bail, P.Y., Billard, R., 1983. Gonadotrophine glycoproteique maturante et
oestradiol-17b pendant le cycle reproducteur chez la truite fario (Salmo trutta) femelle. Gen. Comp. Endo-
crinol. 49, 220 – 231.
Bromage, N., Cumaranatunga, R., 1988. Egg production in the rainbow trout. In: Muir, J.F., Roberts, R.J. (Eds.),
Recent Advances in Aquaculture, vol. 3. Croom Helm, London, pp. 63 – 138.
Bromage, N., Elliot, J., Springate, J.R., Whitehead, C., 1984. The effects of constant photoperiods on the timing
of spawning in the rainbow trout. Aquaculture 43, 213 – 223.
Bromage, N., Hardiman, P., Jones, J., Springate, J., Bye, V., 1990. Fecundity, egg size and total egg volume
differences in 12 stocks of rainbow trout, Oncorhynchus mykiss Richardson. Aquacult. Fish. Manage. 21,
269 – 284.
Bromage, N., Randall, C., Davies, B., Thrush, M., Duston, J., Carillo, M., Zanuy, S., 1993. Photoperiodism
and the control of reproduction and development in farmed fish. In: Lahlou, B., Vitiello, P. (Eds.),
Aquaculture: Fundamental and Applied Research. American Geophysical Union, Washington, DC, pp.
81 – 102.
Buss, K.W., 1982. Photo-period manipulation of the spawning season of brood trout. In: Bulleid, M. (Ed.),
Proceedings of the Institute of Fisheries Management Commercial Trout Farming Symposium. Institute of
Fisheries Management, Reading, pp. 116 – 126.
Chmilevsky, D.A., 1999. Effects of extreme ranged temperatures on fish oogenesis. Proceedings from 6th
International Symposium on the Reproductive Physiology of Fish, July 4 – 9, 1999. University of Bergen,
Bergen, Norway, p. 316.
Craik, J.C.A., Harvey, S.M., 1984. Egg quality in rainbow trout: the relation between egg viability, selected aspect
of egg composition, and time of stripping. Aquaculture 40, 115 – 134.
Davies, B., Bromage, N.R., 1991. The effects of fluctuating seasonal and constant temperatures on the photo-
periodic advancement of reproduction in female rainbow trout. In: Scott, A.P., Sumpter, J.P., Kime, D.E.,
Rolfe, M.S. (Eds.), Proceedings of the 4th International Symposium on Reproductive Physiology of Fish.
University of East Anglia, Norwich, UK, pp. 154 – 156, FishSymp 91, Sheffield.
Dubois, R.B., Plaster, S.D., Rasmussen, P.W., 1989. Fecundity of spring- and fall-run steelhead from 2 western
lake-superior tributaries. Trans. Am. Fish. Soc. 118, 311 – 316.
Duston, J., 1987. Circannual rhythms of maturation in the rainbow trout and their entrainment by changes in
photoperiod. PhD thesis, Aston University, UK.
Duston, J., Bromage, N., 1986. Photoperiodic mechanisms and rhythms of reproduction in the female rainbow
trout. Fish Physiol. Biochem. 2, 35 – 51.
 B. Davies, N. Bromage / Aquaculture 205 (2002) 183–200 199

Duston, J., Bromage, N., 1987. Constant photoperiod regimes and the entrainment of the annual cycle of
reproduction in the female rainbow trout (Salmo gairdneri). Gen. Comp. Endocrinol. 65, 373 – 384.
Duston, J., Bromage, N., 1988. The entrainment and gating of the endogenous circannual rhythm of repro-
duction in the female rainbow trout (Salmon gairdneri). J. Comp. Physiol. 164(A), 259 – 268.
Escaffre, A.M., Billard, R., 1979. Changes in fertilizability of rainbow trout eggs left in the abdominal cavity
during the post-ovulatory period. Bull. Fr. Piscic. 272, 56 – 70.
Escaffre, A.M., Petit, J., Billard, R., 1977. Changes in the yield of eggs and of their fertility at various times after
ovulation in the rainbow trout. Bull. Fr. Piscic. 265, 134 – 142.
Goryczko, K., 1972. A change of the spawning season in rainbow trout (Salmo gairdneri Richardson). Rocz.
Nauk Roln. 94H, 57 – 68.
Henderson, N.E., 1963. Influence of light and temperature on the reproductive cycle of the eastern brook trout,
Salvelinus fontinalis (Mitchell). J. Fish. Res. Board Can. 20, 859 – 897.
Johnstone, C.E., Hambrook, M., Gray, R.W., Davidson, K.G., 1992. Manipulation of reproductive function in
Atlantic salmon (Salmo salar) kelts with controlled photoperiod and temperature. Can. J. Fish. Aquat. Sci.
49, 2055 – 2061.
Kime, D.E., 1979. The effect of temperature on the testicular steroidogenesic enzymes of the rainbow trout,
Salmo gairdneri. Gen. Comp. Endocrinol. 39, 290 – 296.
Kime, D.E., 1980. Androgen biosynthesis by testis of the goldfish Carassius auratus in vitro: the effect of
temperature on the formation of steroid glucuronides. Gen. Comp. Endocrinol. 41, 164 – 172.
King, H.R., Pankhurst, N.W., 1999. Ovulation of Tasmanian Atlantic salmon maintained at elevated temper-
atures: implications of climate change for sustainable industry development. Proceedings from 6th Interna-
tional Symposium on the Reproductive Physiology of Fish, July 4 – 9, 1999. University of Bergen, Bergen,
Norway, pp. 396 – 398.
Korsgaard, B., Mommsen, T.P., Saunders, R.L., 1986. The effect of temperature on the vitellogenic response in
Atlantic salmon post-smolts (Salmo salar). Gen. Comp. Endocrinol. 62, 193 – 201.
Mackay, M.E., Lazier, C.B., 1993. Estrogen responsiveness of vitellogenin gene expression in rainbow trout
(Oncorhynchus mykiss) kept at different temperatures. Gen. Comp. Endocrinol. 89, 255 – 266.
Morrison, J.K., Smith, C.E., 1986. Altering the spawning of rainbow trout by manipulating water temperature.
Prog. Fish-Cult. 48, 52 – 54.
Munro, A.D., 1990. Tropical freshwater fish. In: Munro, A.D., Scott, A.P., Lam, T.J. (Eds.), Reproductive
Seasonality in Teleosts: Environmental Influences. CRC Press, Boca Raton, FL, pp. 145 – 239.
Nakari, T., Soivio, A., Pesonen, S., 1987. Effects of an advanced photoperiod cycle on the gonadal development
and spawning time of 2-year-old Salmo gairdneri R. reared in earth ponds under extreme annual water
temperatures. Aquaculture 67, 369 – 384.
Nakari, T., Soivio, A., Pesonen, S., 1988. The ovarian development and spawning time of Salmo gairdneri R.
reared in advanced and delayed annual photoperiod cycles at naturally fluctuating water temperatures in
Finland. Ann. Zool. Fenn. 25, 335 – 340.
Nomura, M., 1962. Studies on reproduction of rainbow trout, Salmo gairdneri, with special reference to egg
taking: III. Acceleration of spawning by control of light. Bull. Jpn. Soc. Sci. Fish. 28, 1070 – 1076.
Nomura, M., 1963. Studies on reproduction of rainbow trout, Salmo gairdneri, with special reference to egg
taking-IV. The fecundity or number and weight of eggs taken. Bull. Jpn. Soc. Sci. Fish. 29, 325 – 333.
Nomura, M., Sakai, K., Takashima, F., 1974. The over-ripening phenomenon of rainbow trout: I. Temporal mor-
phological changes of eggs retained in the body cavity after ovulation. Bull. Jpn. Soc. Sci. Fish. 40, 977 – 984.
Olin, T., von der Decken, A., 1989. Vitellogenin synthesis in Atlantic salmon (Salmo salar) at different
acclimation temperatures. Aquaculture 79, 397 – 402.
Pankhurst, N.W., Thomas, P.M., 1998. Maintenance at elevated temperature delays the steroidogenic and
ovulatory responsiveness of rainbow trout Oncorhynchus mykiss to luteinizing hormone releasing hormone
analogue. Aquaculture 166, 163 – 177.
Pankhurst, N.W., Purser, G.J., Van Der Kraak, G., Thomas, P.M., Forteath, G.N.R., 1996. Effect of holding
temperature on ovulation, egg fertility, plasma levels of reproductive hormones and in vitro ovarian steroido-
genesis in the rainbow trout Oncorhynchus mykiss. Aquaculture 146, 278 – 290.
Randall, C.F., 1992. Photoperiodic control of reproduction and patterns of melatonin secretion in the rainbow
trout, Oncorhynchus mykiss. PhD Thesis, Stirling University, UK.
200 B. Davies, N. Bromage / Aquaculture 205 (2002) 183–200

Sakai, K., Nomura, M., Takashima, F., Oto, H., 1975. The over-ripening phenomenon of rainbow trout: II.
Changes in the percentage of eyed eggs, hatching rate and incidence of abnormal alvins during the process
of over-ripening. Bull. Jpn. Soc. Sci. Fish. 41, 855 – 860.
Sokal, R.R., Rohlf, F.J., 1981. Biometry. W.H. Freeman and Company, San Francisco, 859 pp.
Springate, J.R.C., 1985. Egg quality and fecundity in rainbow trout: the determination factors and mechanisms of
control, PhD Thesis, Aston University, UK.
Springate, J.R.C., Bromage, N.R., 1985. Effects of egg size on early growth and survival in rainbow trout (Salmo
gairdneri Richardson). Aquaculture 47, 163 – 172.
Springate, J.R.C., Bromage, N.R., Elliott, J.A.K., Hudson, D.L., 1984. The timing of ovulation and stripping and
their effects on the rates of fertilization and survival to eying, hatching and swim-up in the rainbow trout
(Salmo gairdneri R.). Aquaculture 43, 313 – 322.
Taranger, G.L., 1993. Sexual maturation in Atlantic salmon, Salmo salar L.; aspects of environmental and
hormonal control, PhD thesis, University of Bergen, Norway.
Taranger, G.L., Hansen, T., 1993. Ovulation and egg survival following exposure of Atlantic salmon, Salmo salar
L., broodstock to different water temperatures. Aquacult. Fish. Manage. 24, 151 – 156.
Taranger, G.L., Haux, C., Stefansson, S.O., Bjornsson, B.Th., Walther, B.Th., Hansen, T., 1998. Abrupt changes
is photoperiod affect age at maturity, timing of ovulation and plasma testosterone and estradiol-17b profiles
in Atlantic salmon, Salmo salar. Aquaculture 162, 85 – 98.
Taranger, G.L., Stefansson, S.O., Oppedal, F., Andersson, E., Hansen, T., Norberg, B., 1999. Photoperiod and
temperature affects gonadal development and spawning time in Atlantic salmon (Salmo salar L.). Proceedings
from 6th International Symposium on the Reproductive Physiology of Fish, July 4 – 9, 1999. University of
Bergen, Bergen, Norway, p. 345.
Titarev, F., 1975. Acceleration of maturation in the rainbow trout (Salmo gairdneri) under the influence of
increased water temperature. J. Ichthyol. 15, 507 – 509.
Tyler, C.R., Sumpter, J.P., Bromage, N.R., 1987. Uptake of vitellogenin into cultured ovarian follicles of
rainbow trout. In: Idler, D.R., Crim, L.W., Walsh, J.M. (Eds.), Proceedings from 3rd International Sympo-
sium on the Reproductive Physiology of Fish. Memorial University of Newfoundland, St. John’s,
Newfoundland, p. 221.