Journal of Fish Biology (2014) 85, 1163–1176

doi:10.1111/jfb.12481, available online at wileyonlinelibrary.com

Osmoregulation in Atlantic salmon Salmo salar smolts

transferred to seawater at different temperatures

S. O. Handeland*†#, A. K. Imsland‡§‖#, T. O. Nilsen‡, L. O. E.

Ebbesson*, C. D. Hosfeld†, C. Pedrosa*, H. Toften¶ and S. O.
Stefansson‡
*Uni Research AS, 5020 Bergen, Norway, †Bergen University College, 5020 Bergen Norway,
‡Department of Biology, University of Bergen, High Technology Centre, 5020 Bergen,
Norway, §Akvaplan-niva Iceland Office, Akralind 4, 201 Kópavogur, Iceland and ¶Nofima,
Muninbakken 9-13 Breivika, 9291 Tromsø, Norway

In order to investigate how changes in gill Na+ , K+ -ATPase (NKA) 𝛼1a and 𝛼1b subunits, Na+ , K+ ,
2Cl− co-transporter (NKCC1) and the apical cystic fibrosis trans-membrane conductance regulator-I
(CFTR-I) transcripts in wild strain of Atlantic salmon, Salmo salar, smolts are affected by temperature
during spring, hatchery-reared parr (mean ± s.e. fork length = 14⋅1 ± 0⋅5; mean ± s.e. body mass =
28⋅5 ± 4⋅5 g) originating from broodstock from the Vosso river (western Norway) were acclimated
to three temperature regimes (4⋅1, 8⋅1 and 12⋅9∘ C) in May and reared under gradually increasing
salinity between May and June. Changes in plasma Na+ , haematocrit (Hct) and PCO2 were monitored
in order to assess and compare key physiological changes with the transcriptional changes in key
ion transporters. The temperatures reflect the natural temperature range in the River Vosso during
late spring. Overall, higher gill NKA 𝛼1b mRNA levels, gill NKCC1a levels and CFTR-I levels were
observed in the 4⋅1∘ C group compared to the 11⋅9∘ C group. This coincided with a 2–3 week period
with decreased Hct and PCO2 and may indicate a critical window when smolts suffer from reduced
physical performance during migration. Further research is needed to confirm the potential interaction
between ecological and physiological conditions on mortality of hatchery-reared smolts from River
Vosso during their natural migration.
© 2014 The Fisheries Society of the British Isles

Key words: environmental manipulation; smoltification sea migration; wild strain.

INTRODUCTION
Atlantic salmon Salmo salar L. 1758 exhibits a life-history pattern with spawning
and juvenile stages in fresh water (FW), followed by parr–smolt transformation that
pre-adapts the juvenile S. salar for a transition and migration into the marine envi-
ronment (McCormick & Saunders, 1987; Stefansson et al., 2008; McCormick, 2013).
The parr–smolt transformation takes place during increasing day length in spring and
the single most critical part involved in this biological process is the development of
hypo-osmoregulatory ability (McCormick & Saunders, 1987; Stefansson et al., 2008).

‖Author to whom correspondence should be addressed. Tel: +354 562 58 00; email: ai@akvaplan.niva.no
#Authors making an equal contribution

1163
© 2014 The Fisheries Society of the British Isles
1164 S . O . H A N D E L A N D E T A L.

Increasing hypo-osmoregulatory ability is the result of several physiological and

anatomical changes in the gut, kidney and gills, including the differentiation of the FW
mitochondria-rich chloride cells (CCs) to the seawater (SW) CC type with elevated
gill Na+ , K+ -ATPase activity (gill NKA; Borgattti et al., 1992; McCormick, 2013).
In FW, the basolateral gill NKA is known to be the primary driving force for uptake of
monovalent ions (Na+ and Cl− ) in combination with apical Na+ channels and Na+ /H+
and Cl− /HCO3 − exchangers (Evans et al., 2005). NKA works in conjunction with the
H+ -ATPase (Avella & Bornancin, 1989). In SW-acclimated smolts, the same enzyme
stimulates the excretion of excess Na+ and Cl− in concert with ion transporters such
as Na+ , K+ , 2Cl− co-transporter (NKCC1a) and the apical cystic fibrosis transmem-
brane conductance regulator (CFTR-I). Gill NKA consists of three subunits, with the
𝛼-subunit containing the active binding site. The differential expression of 𝛼1a and 𝛼1b
isoforms during smolting further suggests a mechanism by which the S. salar modu-
lates gill NKA as a pre-adaptation to altered salinity during migration (Richards et al.,
2003; Bystriansky et al., 2006; Nilsen et al., 2007; Madsen et al., 2009; Stefansson
et al., 2012).
In wild S. salar, the first weeks of marine life represent a critical period and are occa-
sionally characterized by high mortality caused by both osmoregulatory challenges and
exposure to marine predators. Following SW transfer, plasma ion levels increase tran-
siently (Stagg et al., 1989; Handeland et al., 1996) with a concurrent tissue dehydra-
tion (Blackburn & Clarke, 1987; McCormick et al., 1989; Handeland et al., 1996). In
S. salar smolts acclimated to SW at 4⋅3∘ C compared to 9⋅4∘ C, the physiological dis-
turbance was greater and the recovery was slower at low temperature (Handeland et al.,
2003). This suggests a temperature-dependent response in smolts ability to adapt to SW.
The main smolt migration in the River Vosso (Hordaland, western Norway) normally
takes place from 10 May onwards, with a 50% accumulative migration rate between
15 and 20 May (Barlaup, 2008). The time from when migrants leave the River Vosso,
until they pass Salhus (the end of the Osterfjord system), may take 2 weeks and during
this passage salinity varies from 2 to 30 at the surface layer.
The main objective of this study has been to clarify how SW temperature
affects the intensity and duration of osmotic disturbances during SW acclimation
of hatchery-reared smolts from the Vosso strain. Short and long-term changes in gill
NKA 𝛼1a, 𝛼1b, NKCC1a and CFTR-I transcripts and gill NKA activity in groups of
smolts exposed to SW at 4⋅1, 8⋅1 and 11⋅9∘ C over a period of 30 days were studied. In
addition, changes in plasma Na+ , haematocrit (Hct) and PCO2 were monitored in order
to assess and compare key physiological changes with the transcriptional changes in
key ion transporters and to what extent those changes are affected by temperature.

MATERIALS AND METHODS

THE FISH STOCKS AND REARING CONDITIONS

The fish used were 1+ year S. salar smolts of the Vosso strain. The eggs hatched in mid-April
2009 and the fry were first fed at the end of May. From first feeding until the end of August,
the fish were kept on a simulated natural photoperiod for the Vosso area (60∘ 30′ N; 6∘ 3′ E),
ambient water temperature and were fed a standard dry diet (Ewos; www.ewos.com) with ration
according to recommendation based on temperature and fish size (Austreng et al., 1987). On
1 September, all fish were transferred from the Voss hatchery to Lake Evanger and reared in a net

© 2014 The Fisheries Society of the British Isles, Journal of Fish Biology 2014, 85, 1163–1176
 S M O LT I F I C AT I O N I N A W I L D S T R A I N O F S A L M O S A L A R 1165

pen (20 m × 20 m, depth 5 m). On 28 December, c. 650 pre-smolts were randomly collected and
brought from Lake Evanger to the Industrial and Aquatic Laboratory (ILAB) at the Bergen High
Technology Centre. After arrival at ILAB, the fish were randomly distributed into three 1 m2
square fibre-glass tanks with a rearing volume of 500 l (n = 212 pr tank). Each tank was initially
supplied with running FW (25 l min−1 , temperature: 8⋅1∘ C). A simulated natural photoperiod
(60∘ 24′ N; 6∘ 3′ E) was provided by a fluorescent daylight tube mounted under the tank cover.
All groups were fed in excess with the same dry diet from automatic feeders during light hours
(Nutra svev, Skretting AS; www.skretting.no). Temperature and oxygen content were recorded
daily and any dead fish were removed immediately.

SW EXPERIMENT
On 20 May, the smolts showed elevated gill NKA (mean ± s.e. = 11⋅0 ± 2⋅9, μmol ADP mg
protein−1 h−1 ) and normal morphological signs of smolting, i.e. dark fin margins, absence of parr
marks and loose, silvery scales. The fish [mean ± s.e. fork length (LF ) = 14⋅1 ± 0⋅5; mean ± s.e.
body mass = 28⋅5 ± 4⋅5 g] were randomly distributed into 6 1 m2 rearing tanks. On 21 May
over a period of 1 week, the temperature was gradually altered from 8⋅1 to 4⋅1∘ C (s.d. = 0⋅4∘
C, two tanks), raised to 11⋅9∘ C (s.d. = 0⋅3∘ C, two tanks), or kept at 8⋅1∘ C (s.d. = 0⋅4∘ C, two
tanks). Starting on 26 May, salinity in all tanks was increased on a weekly basis. The salinity was
increased from FW to 14⋅3 between 26 May and 2 June, from 14⋅3 to 25⋅1 between 2 and 8 June
and from 25⋅1 to 33⋅7 from 8 June onwards. For each period, the desired salinity was achieved
by progressively mixing SW with FW, increasing salinity at a rate of 2–3 h−1 . Temperature,
salinity and oxygen content were recorded daily. In SW, the fish were fed in excess with a
commercial dry diet (Respons, Skretting AS) from automatic feeders during daylight between
0900 and 1600 hours. Water flow was kept at 30 l min−1 , maintaining oxygen saturation in the
outlet water above 80%. All tanks were checked daily and dead fish were removed immediately.

S A M P L I N G P RO C E D U R E S
Prior to transfer, on 20 May, and at regular intervals in SW, i.e. 26 May, 2 June, 8 June, 16 June,
23 June and 30 June, random samples of 12 fish were taken from each group. All groups were
starved 24 h prior to sampling and the fish were killed by a blow to the head. From each fish,
blood was drawn into a heparinized syringe from the caudal peduncle. The blood was analysed
immediately using an ISTAT analyser (Abbot Norway AS; www.abottpointofcare.com) with
EC8+ cassettes for plasma Na+ , PCO2 and Hct. Plasma PCO2 was adjusted to the temperature
difference between 37∘ C and the temperature of the fish according to Heisler (1984). At the same
dates, gill filaments were sampled, frozen in sucrose-EDTA-imidazole (SEI) buffer, stored on
RNAlater or dry frozen at −80∘ C. Samples in SEI buffer were subsequently analysed for gill
NKA activity with a kinetic assay run in 96 well microplates at 25∘ C and read at a wavelength
of 340 nm for 10 min as described in the study of McCormick (1993). The LF (to nearest 0⋅1
cm) and mass (to nearest 0⋅1 g) of the smolts in each group were determined during sampling.

R N A I S O L AT I O N , CD N A S Y N T H E S I S A N D R E A L- T I M E
Q U A N T I TAT I V E P C R
Total RNA was isolated from c. 50 mg gill tissue by phenol–chloroform extraction using
TRI Reagent (Sigma; www.sigmaaldrich.com) as outlined by Chomczynski (1993). Total
RNA concentration and purity were determined by the NanoDrop ND-1000 UV-Vis spec-
trophotometer (NanoDrop Technologies; www.nanodrop.com) and the RNA integrity was
evaluated with the Agilent 2100 Bioanalyser using the RNA 6000 Nano LabChip kit (Agilent
Technologies; www.agilient.com). Total RNA was treated with RQ1 RNase-free DNase
(Promega; www.promega.com) and cDNA was reversely transcribed using 350 ng total RNA
and random nonamers in conjunction with the reverse transcription core kit (EUROGENTEC
RT-RTCK-05; www.eurogenetic.com) following the manufacturer’s instructions. Gill NKA
𝛼-subunit isoforms (𝛼1a and 𝛼1b), NKCC1a, CFTR-I and elongation factor 1𝛼 (EF1𝛼) mRNA
levels were measured by quantitative polymerase chain reaction (qPCR) assays according to

© 2014 The Fisheries Society of the British Isles, Journal of Fish Biology 2014, 85, 1163–1176
1166 S . O . H A N D E L A N D E T A L.

Nilsen et al. (2007, 2010) using the ABI prism 7000 detection system platform (ABI; Applied
Biosystems; www.appliedbiosystems.com.). Melt-curve analysis verified that the primer sets
for each qPCR assay generated one single product and no primer–dimer artefacts. Primers and
PCR conditions are published by Nilsen et al (2007, 2010). Briefly, for each assay, triplicate
five-fold cDNA dilution series made from total RNA from different exposure groups were used
to determine amplification efficiencies (E) calculated as the slope (b) from the plot of log10 RNA
concentration and threshold cycle (Ct) values using the following formula: E = 10(−1b ) . This
−1

efficiency was used to correct for difference in amplification efficiency when calculating gene
expression according to Pfaffl (2004). Expression is presented as relative to the endogenous
reference gene ef1𝛼 (Olsvik et al., 2005). ef1𝛼 did not change over time or differ between
treatments in this study.

S TAT I S T I C A L A N A LY S I S
Prior to statistical analysis, all data were tested for normality (Kolmogorov–Smirnov test)
and homogeneity of variances among the different groups (Hartly F-max test). Two-way model
III nested ANOVAs (Scheffé, 1959), where replicate tanks (random effect) are nested within
temperatures (fixed effect), were used to test mean values among and within groups at different
sampling dates. The equation of the fully nested model had the form: X ijk = 𝜇 + 𝛼 i + 𝛿 j(i) + 𝜖 ijk ,
where 𝜇 is the general level, 𝛼 i is the treatment effect for temperaturei , 𝛿 j(i) is the replicate factor
(here: tankj ) nested within temperature 𝛼 i and 𝜖 ijk is the model error term.
Significant ANOVAs were followed by a Student–Newman–Keuls (SNK) post hoc test to
locate differences between experimental groups. A linear regression was used to test for differ-
ences over time between experimental groups. A significance level (𝛼) of 0⋅05 was used if not
stated otherwise.

RESULTS

P L A S M A S O D I U M L E V E L S F O L L OW I N G T R A N S F E R TO S W
Following transfer to SW, plasma sodium (Na+ ) levels were significantly influenced
by time and temperature (P < 0⋅001; Fig. 1). In all groups, a significant rise in plasma
Na+ levels from c. 145 to 161, 155 and 157 mmol l−1 was observed in the 4⋅1, 8⋅1 and
11⋅9∘ C groups, respectively (linear regression, P < 0⋅001). During the same period,
the overall plasma Na+ levels were higher in the 4⋅1∘ C group, compared to both the
8⋅1 and the 11⋅9∘ C groups, and levels were significantly higher on 16 June and 23 June
(SNK test, P < 0⋅05).

P L A S M A HC T F O L L O W I N G T R A N S F E R T O S E AWAT E R
A gradual transient decrease in Hct was recorded in all groups between 26 May and 8
June (linear regression, P < 0⋅05; Fig. 2), followed by an overall increase from 16 June
onwards (linear regression, P < 0⋅05). Throughout the first period in SW, between 26
May and 2 June, Hct levels in the 8⋅1∘ C group were significantly higher than in the
other groups (SNK test, P < 0⋅05). On 30 June, after 14 days at 33, Hct levels stabilized
at 29, 31 and 29% in the 4⋅1, 8⋅1 and 11⋅9∘ C groups, respectively.

P L A S M A PC O2 F O L L O W I N G T R A N S F E R T O S E AWAT E R
The measurements of blood PCO2 levels (Fig. 3) followed a similar pattern as
reported for Hct. PCO2 levels were significantly affected by time and temperature

© 2014 The Fisheries Society of the British Isles, Journal of Fish Biology 2014, 85, 1163–1176
 S M O LT I F I C AT I O N I N A W I L D S T R A I N O F S A L M O S A L A R 1167

166

164

162 a

160
a
158
Plasma Na (mmol l–1)

156

154 b
b
152 b
b
150

148

146

144

142

140
26 May 2 June 8 June 16 June 23 June 30 June

Fig. 1. Mean ± s.e. plasma Na levels (n = 12) in Salmo salar smolts during parr–smolt transformation
at 4⋅1 ( ), 8⋅1 ( ) and 11⋅9∘ C ( ). Different lowercase letters indicate significant differences
(Student–Newman–Keuls, P < 0⋅05) between temperature groups at same time of sampling.

(P < 0⋅01). In all groups, a reduction in PCO2 was evident between 26 May and 16
June (linear regression, P < 0⋅05), corresponding to the gradual increase in salinity.
The recorded PCO2 levels in the 8⋅1∘ C group were significantly higher than those
observed in the 11⋅9∘ C group on 2 June, 8 June and 23 June (SNK test, P < 0⋅05).

G I L L N K A F O L L O W I N G T R A N S F E R T O S E AWAT E R
Overall, gill NKA was significantly affected by time and temperature (P < 0⋅001;
Fig. 4). In the 4⋅1∘ C group, the gill NKA remained relatively stable at c. 9 μmol
ADP mg−1 protein h−1 until 16 June. Thereafter, a gradual increase (linear regression,
P < 0⋅05) in gill NKA was observed, reaching peak level at 13 μmol ADP mg−1 pro-
tein h−1 on 30 June. In the 8⋅1∘ C group, gill NKA increased between 8 June and 23
June (linear regression, P < 0⋅001), while in the 11⋅9∘ C group the gill NKA increased
between 2 June and 16 June (linear regression, P < 0⋅001). On the last sampling, 30
June, the highest gill NKA was observed in the 8⋅1 and 11⋅9∘ C groups, significantly
higher than that found in the 4⋅1∘ C group (SNK test, P < 0⋅05).

G I L L N K A 𝛼1A A N D 𝛼1B MR N A L E V E L S F O L L OW I N G T R A N S F E R
T O S E AWAT E R
The development of gill NKA 𝛼1a mRNA levels was significantly influenced by time
(linear regression, P < 0⋅001), whereas no differences were observed among the differ-
ent temperature groups (Fig. 5). A significant decline (linear regression, P < 0⋅001) in

© 2014 The Fisheries Society of the British Isles, Journal of Fish Biology 2014, 85, 1163–1176
1168 S . O . H A N D E L A N D E T A L.

36
35
34 a
33
32
31 a
30
b
Haematocrit (%)

29
28
b
27 b
26
25 b
24
23
22
21
20
26 May 2 June 8 June 16 June 23 June 30 June

Fig. 2. Mean ± s.e. plasma haematocrit levels (n = 12) in Salmo salar smolts during parr–smolt transforma-
tion at 4⋅1 ( ), 8⋅1 ( ) and 11⋅9∘ C ( ). Different lowercase letters indicate significant differences
(Student–Newman–Keuls, P < 0⋅05) between temperature groups at same time of sampling.

gill NKA 𝛼1a mRNA levels was observed in all groups between 26 May and 2 June
reaching transcript levels close to 0 from 8 June onwards. Gill NKA 𝛼1b mRNA lev-
els were significantly affected by time (P < 0⋅001) and temperature (P < 0⋅05; Fig. 6).
Overall, significantly higher gill NKA 𝛼1b mRNA levels were observed in the 4⋅1∘ C
group compared to the 11⋅9∘ C group (SNK test, P < 0⋅05). In all groups, a signif-
icant increase in gill NKA 𝛼1b mRNA levels was observed between 26 May and 16
June (P < 0⋅05; Fig. 6). There were no significant differences among groups during this
period. On 23 June, however, the NKA 𝛼1b mRNA transcript level in the 4⋅1∘ C group
was significantly higher than in the 11⋅9∘ C group (SNK test, P < 0⋅05). Between 16
and 30 June, a further significant rise in NKA 𝛼1b mRNA level from 13⋅7 to 28⋅0 was
evident in the 4⋅1∘ C group (linear regression, P < 0⋅001), whereas no changes were
observed in the other groups. There were no differences among groups upon termina-
tion on 30 June.

G I L L N K C C 1A A N D C F T R- I MR N A L E V E L S F O L L OW I N G
T R A N S F E R T O S E AWAT E R
There was an overall effect of both time and temperature on gill NKCC1a mRNA
and gill CFTR-I mRNA levels (P < 0⋅05; Figs 7 and 8). In line with gill NKA 𝛼1b
mRNA levels, higher gill NKCC1a and CFTR-I mRNA levels were observed in the
4⋅1∘ C group compared to the 11⋅9∘ C group (SNK test, P < 0⋅05). In the 8⋅1∘ C group,
a significant rise in gill NKCC1a mRNA levels was observed between 26 May and 16
June (linear regression, P < 0⋅05; Fig. 7). A similar rise in gill NKCC1a mRNA levels

© 2014 The Fisheries Society of the British Isles, Journal of Fish Biology 2014, 85, 1163–1176
 S M O LT I F I C AT I O N I N A W I L D S T R A I N O F S A L M O S A L A R 1169

26

a
25

24 a
b
a
Blood PCO2 (mmHg)

23

a
22
c a

21 b

20
b

19

18
26 May 2 June 8 June 16 June 23 June 30 June

Fig. 3. Mean ± s.e. plasma partial pressure of carbon dioxide (PCO2 ) (n = 12) in Salmo salar smolts during
parr–smolt transformation at 4⋅1 ( ), 8⋅1 ( ) and 11⋅9∘ C ( ). Different lowercase letters indicate significant
differences (Student–Newman–Keuls, P < 0⋅05) between temperature groups at same time of sampling.

was evident between 26 May and 23 June in the 4⋅1 and 11⋅9∘ C groups (linear regres-
sion, P < 0⋅05). Changes in CFTR-I reflected the observed changes in NKCC1a. In
the 4⋅1 and 8⋅1∘ C groups, transcription levels increased significantly between 26 May
and 30 June, and between 26 May and 23 June, respectively. In contrast, no changes
in CFTR-I mRNA levels were recorded in the 11⋅9∘ C group during the same period.
Consequently, on 30 June, the observed CFTR-I mRNA level in the 4⋅1∘ C group was
significantly higher than in the 11⋅9∘ C group (SNK test, P < 0⋅05). No differences
were observed between the 4⋅1 and 8⋅1∘ C groups at termination on 30 June.

DISCUSSION
In this study, overall mean higher gill NKA 𝛼1b mRNA levels, gill NKCC1a levels
and CFTR-I levels were observed in the 4⋅1∘ C group compared to the 11⋅9∘ C group.
These results may suggest that increased hypo-osmoregulatory capacity in SW, here
mediated as changes in transcriptional and translational levels of basic ion transporters
in S. salar gills, is controlled by temperature.
In S. salar, the gill NKA enzyme is widely accepted to be the driving force for active
excretion of monovalent ions. NKA works in conjunction with the H+ -ATPase (Avella
& Bornancin, 1989). Which one is the primary driving force in salmonids is a matter of
some contention (Parks et al., 2008). To remove the excess of incoming Na+ and Cl−
ions in blood, the gill NKA enzyme pumps three Na+ ions out and two K+ ions into the

© 2014 The Fisheries Society of the British Isles, Journal of Fish Biology 2014, 85, 1163–1176
1170 S . O . H A N D E L A N D E T A L.

28
27
26
25
a
24
23 a
a
22
Gill Na+, K+-ATPase activity

21
20
19
18
17
a
16 b
15
14 b
a
13
b
12 c
11
10 b
c
9
b
8
7
6
26 May 2 June 8 June 16 June 23 June 30 June

Fig. 4. Mean ± s.e. gill Na+ , K+ -ATPase (NKA) activity (n = 12) in Salmo salar smolts during parr–smolt trans-
formation at 4⋅1 ( ), 8⋅1 ( ) and 11⋅9∘ C ( ). Different lowercase letters indicate significant differences
(Student–Newman–Keuls, P < 0⋅05) between temperature groups at same time of sampling.

CCs to set up a Na+ gradient (Evans et al., 2005). The Na+ gradient between the CC and
blood is used to bring Cl− into the cell via the basolateral NKCC1a (Nilsen et al., 2007).
The Cl− then leaves the CC through an apical chloride channel, the CFTR-I. The gill
NKA enzyme consists of different subunits, (𝛼, 𝛽 and 𝛾), where the 𝛼-subunit is con-
sidered to be the major catalytic unit containing all the binding sites for ATP. Recently,
Madsen et al. (2009) and McCormick (2013) showed that the different isoforms of gill
NKA 𝛼-subunit mRNA and protein are primarily located in different FW and SW-type
CCs in S. salar. This general model (Madsen et al., 2009; McCormick, 2013) corre-
sponds to the time line in the present experiment, showing that gill NKA 𝛼1a mRNA
levels were significantly down-regulated in all groups between 26 May and 2 June,
concurrent with a significant up-regulation of the gill NKA 𝛼1b mRNA levels between
26 May and 16 June. The present results are in line with previous studies on rainbow
trout Oncorhynchus mykiss (Walbaum 1792) (Richards et al., 2003) and S. salar (Bys-
triansky et al., 2006; Nilsen et al., 2007), confirming that S. salar can modulate the
expression of 𝛼-subunit isoforms in response to changes in salinity. McCormick (2013)
provides evidence that changes in transcriptional levels parallel consistent changes in
protein abundance, and that the two isoforms have different physiological function in
S. salar smolts. Consequently, the rapid down-regulation of gill NKA 𝛼1a mRNA lev-
els indicates that this FW CC isoform is less important in saline water. On the contrary,
the elevated expression pattern of gill NKA 𝛼1b mRNA upon SW exposure suggests
that the corresponding SW CC ion-regulatory protein is essential for ion secretion in
post-smolts.

© 2014 The Fisheries Society of the British Isles, Journal of Fish Biology 2014, 85, 1163–1176
 S M O LT I F I C AT I O N I N A W I L D S T R A I N O F S A L M O S A L A R 1171

3
Gill NKA 1a mRNA

–1
26 May 2 June 8 June 16 June 23 June 30 June

Fig. 5. Mean ± s.e. gill Na+ , K+ -ATPase (NKA) 𝛼1a mRNA expression (n = 12) in Salmo salar smolts during
parr–smolt transformation at 4⋅1 ( ), 8⋅1 ( ) and 11⋅9∘ C ( ).

Throughout this study, a continuous increase in transcription of gill NKCC1a and

CFTR-I mRNA levels was observed in the 4⋅1 and 8⋅1∘ C groups. During the same
period, no changes in CFTR-I mRNA levels were recorded in the 11⋅9∘ C group.
Overall, the present findings correspond to the results of Singer et al. (2002), showing
an increase in NKCC1a and CFTR-I mRNA levels following transfer to SW, sug-
gesting both transporters to be important in the activation of CC at higher salinities.
Interestingly, in all groups, a significant up-regulation of gill NKA 𝛼1b mRNA levels
was evident during the first 3 weeks in SW (between 26 May and 16 June), in contrast
to the gill NKA activity in the 4⋅1∘ C group that remained at a relatively low and stable
level until 16 June. Thereafter, a gradual increase in gill NKA was observed, reaching
peak level on 30 June. A similar time lag between elevated gill NKA 𝛼1b mRNA level
and the increase in gill NKA activity has been reported by Seidelin et al. (2000) when
salmonids are transferred from FW to SW. According to Seidelin et al. (2000), this
apparent lag in response time may be explained as a delay of the protein synthesis
compared with gene expression. Accordingly, in the 8⋅1 and 11⋅9∘ C groups, gill NKA
increased on 8 and 2 June respectively, suggesting that the delay is shorter at higher
temperatures.
The present data show a significant temperature effect on development of plasma
sodium levels following transfer to SW, i.e. significantly elevated plasma sodium
levels were observed in the 4⋅1 group compared to the 8⋅1 and 11⋅9∘ C groups on
16 and 23 June. The combination of low NKA activity and elevated plasma sodium
level at 4⋅1∘ C may suggest osmoregulatory perturbation at this temperature and, that
the capacity to excrete ions and reduce fluxes is temporarily reduced. This possible
osmoregulatory perturbation corresponds to the results of Sigholt & Finstad (1990)

© 2014 The Fisheries Society of the British Isles, Journal of Fish Biology 2014, 85, 1163–1176
1172 S . O . H A N D E L A N D E T A L.

40

35

30
Gill NKA 1b mRNA

25 a

20 ab

15
b
10

0
26 May 2 June 8 June 16 June 23 June 30 June

Fig. 6. Mean ± s.e. gill Na+ , K+ -ATPase (NKA) 𝛼1b mRNA expression (n = 12) in Salmo salar smolts during
parr–smolt transformation at 4⋅1 ( ), 8⋅1 ( ) and 11⋅9∘ C ( ). Different lowercase letters indicate significant
differences (Student–Newman–Keuls, P < 0⋅05) between temperature groups at same time of sampling.

and Handeland et al. (2000), who reported increased osmotic disturbances in S. salar
smolts transferred to SW at low temperatures. It should be noted, however, that this
is not a universal finding in all salinities, as S. salar juveniles acclimated to low
ambient temperature in FW had higher gill NKA relative to fish acclimated to higher
temperatures during winter (McCormick et al., 2000; Handeland et al., 2004). In
salmonids, the effect of low temperatures on ion regulation is considered to be caused
by different temperature effects on the ion excretion and the leakage pathway in gills,
i.e. the ion pump generally has a higher Q10 than the leaks (Krumschnabel et al.,
2000). Low temperatures may also cause a depression in the rate of ATP synthesis,
which in turn may be greater than the reduced demand for gill NKA pumping activity.
Taken together, it is suggested that the elevated plasma sodium level in the 4⋅1∘ C
group in this study, compared to the levels in the 8⋅1 and 11⋅9∘ C groups, is linked
with the observed reduced NKA activity in this group.
In all groups, a transient overall decrease in Hct levels between 26 May and 16 June
was evident concurrent with an observed rise in plasma sodium levels. The decrease
in Hct levels may be explained in terms of red blood cell shrinkage caused by high
plasma ion levels, increasing the osmotic gradient across the erythrocyte membrane
(Bath & Eddy, 1979). In O. mykiss, low Hct levels are associated with decreased
oxygen-carrying capacity of blood (Nikinmaa & Soivio, 1979), whereas increased ion
concentrations have been shown to decrease hemoglobin-oxygen affinity in sockeye
salmon Oncorhynchus nerka (Walbaum 1792) (Sauer & Harrington, 1988). During
SW transfers of S. salar smolts, Stagg et al. (1989) reported a transient decrease in

© 2014 The Fisheries Society of the British Isles, Journal of Fish Biology 2014, 85, 1163–1176
 S M O LT I F I C AT I O N I N A W I L D S T R A I N O F S A L M O S A L A R 1173

0 ·8

0 ·7

0 ·6
Gill NKCC 1a mRNA

0 ·5

0 ·4

0 ·3

0 ·2

0 ·1

0 ·0
26 May 2 June 8 June 16 June 23 June 30 June

Fig. 7. Mean ± s.e. gill Na+ , K+ , 2Cl− cotransporter (NKCC1a) mRNA expression (n = 12) in S. salar smolts
during parr–smolt transformation at 4⋅1 ( ), 8⋅1 ( ) and 11⋅9∘ C ( ).

arterial blood oxygen tension (Pa O2 ), returning to control levels within 24 h. The
previously proposed effect of osmotic perturbation on Hct values (Bath & Eddy, 1979),
Pa O2 levels and blood-oxygen content (Stagg et al., 1989) following SW transfer may
help to explain the overall, and corresponding, decrease in PCO2 in this study. There
is a lack of studies, however, explaining the direct cause of such a reduction in PCO2 .
A decrease in oxygen-carrying capacity is assumed to affect the oxygen concentration
of muscle tissue, thus reducing the swimming capacity of the fish (Heisler, 1980).
This explanation raises a new question: What are the potential physiological effects
of high plasma ion levels, decreased Hct and PCO2 levels during early seaward
migration in S. salar smolts? Further studies are warranted to study this effect in
detail.
During natural migration, wild smolts from the River Vosso migrate 80 km through
a fjord system with salinity ranging from c. 2 to 30 and temperatures ranging from 7
to 10∘ C at the surface layer. Most of the records available show that migrating smolts
make only a limited use of the estuary and that migrants move relatively fast to the
ocean, depending on the speed and directions of the surface current (Finstad et al.,
2005).
Recently, the migration pattern outside the River Vosso was characterized by using
radiotagged hatchery-reared smolts from the Vosso strain (T. O. Skilbrei, unpubl. data).
This tracking experiment showed that mean time from when the smolts were released
at the location Stanghelle (3 km from the estuary) until they entered full strength SW
(Salhus) ranged from 2 to 14 days. During this period, c. 70% of the smolts were lost,
suggesting high mortality from marine and avian predators on migrants. The present

© 2014 The Fisheries Society of the British Isles, Journal of Fish Biology 2014, 85, 1163–1176
1174 S . O . H A N D E L A N D E T A L.

0·06

0·05

a
0·04
Gill CFTR-I mRNA

ab
0·03

b
0·02

0·01

0·00
26 May 2 June 8 June 16 June 23 June 30 June

Fig. 8. Mean ± s.e. cystic fibrosis trans-membrane conductance regulator-I (CFTR-I) mRNA expression (n = 12)
in Salmo salar smolts during parr–smolt transformation at 4⋅1 ( ), 8⋅1 ( ) and 11⋅9∘ C ( ). Different
lowercase letters indicate significant differences (Student–Newman–Keuls, P < 0⋅05) between temperature
groups at same time of sampling.

results, indicating a 2–3 week period with decreased Hct and PCO2 levels in combi-
nation with osmoregulatory challenges especially at low temperatures (4⋅1∘ C group),
may indicate a critical window during migration. Further research is needed to confirm
the potential interaction between ecological and physiological conditions on mortality
of hatchery-reared smolts from the River Vosso.

We thank the staff at ILAB for their assistance during this experiment. This study was financed
by grant from the Norwegian Research Council. The experiment described has been approved
by the local responsible laboratory animal science specialist under the surveillance of the Nor-
wegian Animal Research Authority (NARA) and registered by the Authority.

References
Austreng, E., Storebakken, T. & Åsgård, T. (1987). Growth rate estimates for cultured Atlantic
salmon and rainbow trout. Aquaculture 60, 157–160.
Avella, M. & Bornancin, M. (1989). A new analysis of ammonia and sodium transport through
the gills of the fresh-water rainbow-trout (Salmo gairdneri). Journal of Experimental
Biology 142, 155–175.
Barlaup, B. T. (2008). Utredning 2008-9. Nå eller aldri for Vossolaksen- anbefalte tiltak med
bakgrunn i bestandsutvikling og trusselfaktorer. Trondheim: Direktoratet for natturfor-
valtning.
Bath, R. N. & Eddy, F. B. (1979). Ionic and respiratory regulation in rainbow-trout during rapid
transfer to seawater. Journal of Comparative Physiology 134, 351–357.
Blackburn, J. & Clarke, W. C. (1987). Revised procedure for the 24 hour seawater challenge
test to measure seawater adaptability of juvenile salmonids. Canadian Technical Report
of Fisheries and Aquatic Sciences 1515, 1–35.

© 2014 The Fisheries Society of the British Isles, Journal of Fish Biology 2014, 85, 1163–1176
 S M O LT I F I C AT I O N I N A W I L D S T R A I N O F S A L M O S A L A R 1175

Borgattti, A. R., Pagliarani, A. & Ventrella, V. (1992). Gill (Na+ + K+ )-ATPase involvement
and regulation during salmonid adaptation to salt water. Comparative Biochemistry and
Physiology 102A, 637–643.
Bystriansky, J. S., Richards, J. G., Schulte, P. M. & Ballantyne, J. S. (2006). Reciprocal expres-
sion of gill Na+ /K+ -ATPase 𝛼-subunit isoforms 𝛼-1a and 𝛼-1b during seawater acclima-
tion of three salmonid fishes that vary in their salinity tolerance. Journal of Experimental
Biology 209, 1848–1858.
Chomczynski, P. (1993). A reagent for the single-step simultaneous isolation of RNA, DNA and
proteins from cell and tissue samples. Biotechniques 15, 532.
Evans, D. H., Piermarini, P. M. & Choe, K. P. (2005). The multifunctional fish gill: dominant
site of gas exchange, osmoregulation, acid-base regulation, and excretion of nitrogenous
waste. Physiological Reviews 85, 97–177.
Finstad, B., Okland, F., Thorstad, E. B., Bjorn, P. A. & McKinley, R. S. (2005). Migration
of hatchery-reared Atlantic salmon and wild anadromous brown trout post-smolts in a
Norwegian fjord system. Journal of Fish Biology 66, 86–96.
Handeland, S. O., Jarvi, T., Ferno, A. & Stefansson, S. O. (1996). Osmotic stress, antipredator
behaviour, and mortality of Atlantic salmon (Salmo salar) smolts. Canadian Journal of
Fisheries and Aquatic Sciences 53, 2673–2680.
Handeland, S. O., Berge, Å., Björnsson, B. Th., Lie, Ø. & Stefansson, S. O. (2000). Seawater
adaptation of Atlantic salmon (Salmo salar L.) smolts at different temperatures. Aqua-
culture 181, 377–396.
Handeland, S. O., Björnsson, B. Th., Arnesen, A. M. & Stefansson, S. O. (2003). Seawater
adaptation and growth of post-smolt Atlantic salmon (Salmo salar L.) of wild and farmed
strains. Aquaculture 220, 367–384.
Handeland, S. O., Wilkinson, E., Sveinsbø, B., McCormick, S. D. & Stefansson, S. O. (2004).
Temperature influence on the development and loss of seawater tolerance in two
fast-growing strains of Atlantic salmon. Aquaculture 233, 513–529.
Heisler, N. (1980). Role of ion transfer processes in acid-base regulation with temperature-
changes in fish. American Journal of Physiology 264, R441–R451.
Heisler, N. (1984). Acid-base regulations in fishes. In Fish Physiology, Vol. XA (Hoar, W. S. &
Randall, D. J., eds), pp. 315–401. New York, NY: Academic Press.
Krumschnabel, G., Schwarzbaum, P. J., Lisch, J., Biasi, C. & Wieser, W. (2000). Oxygen-
dependent energetics of anoxia-tolerant and anoxia-intolerant hepatocytes. Journal of
Experimental Biology 203, 951–959.
Madsen, S. S., Kiilerich, P. & Tipsmark, C. K. (2009). Multiplicity of expression of
Na+,K+–ATPase 𝛼-subunit isoforms in the gill of Atlantic salmon (Salmo salar):
cellular localisation and absolute quantification in response to salinity change. Journal
of Experimental Biology 212, 78–88.
McCormick, S. D. (1993). Methods for nonlethal gill biopsy and measurements of
Na+ ,K+ –ATPase activity. Canadian Journal of Fisheries and Aquatic Sciences 50,
656–658.
McCormick, S. D. (2013). Smolt physiology and endocrinology. In Fish Physiology, Vol. 32
(McCormick, S. D., Farrel, A. P. & Brauner, C. J., eds), pp. 199–251. Amsterdam: Aca-
demic Press.
McCormick, S. D. & Saunders, R. L. (1987). Preparatory physiological adaptations for marine
life of salmonids: osmoregulation, growth, and metabolism. American Fisheries Society
Symposium 1, 211–229.
McCormick, S. D., Moyes, C. D. & Ballantyne, J. S. (1989). Influence of salinity on the energet-
ics of gill and kidney of Atlantic salmon (Salmo salar). Fish Physiology and Biochemistry
6, 243–254.
McCormick, S. D., Moriyama, S. & Björnsson, B. Th. (2000). Low temperature limits photope-
riod control of smolting in Atlantic salmon through endocrine mechanisms. American
Journal of Physiology 278, 1352–1361.
Nikinmaa, M. & Soivio, A. (1979). Oxygen dissociation curves and oxygen capacities of blood
of a freshwater-fish, Salmo gairdneri. Annales Zoologici Fennici 16, 217–221.
Nilsen, T. O., Ebbesson, L. O. E., Madsen, S. S., McCormick, S. D., Andersson, E., Björns-
son, B. Th., Prunet, P. & Stefansson, S. O. (2007). Differential expression of gill

© 2014 The Fisheries Society of the British Isles, Journal of Fish Biology 2014, 85, 1163–1176
1176 S . O . H A N D E L A N D E T A L.

Na+ ,K+ -ATPase 𝛼- and 𝛽-subunits, Na+ ,K+ ,2Cl – cotransporter and CFTR anion chan-
nel in juvenile anadromous and landlocked Atlantic salmon Salmo salar. Journal of
Experimental Biology 210, 2885–2896.
Nilsen, T. O., Ebbesson, L. O. E., Kverneland, O. G., Kroglund, F., Finstad, B. & Stefansson, S.
O. (2010). Effects of acidic water and aluminium exposure on gill Na+, K+ ATPase – 𝛼
subunit isoforms, enzyme activity, physiology and return rates in Atlantic salmon (Salmo
salar L.). Aquatic Toxicology 97, 250–259.
Olsvik, P. A., Lie, K. K., Jordal, A. E. O., Nilsen, T. O. & Hordvik, I. (2005). Evaluation of
potential reference genes in real-time RT-PCR studies of Atlantic salmon. BMC Molec-
ular Biology 6, 21.
Parks, S. K., Tresguerres, M. & Goss, G. G. (2008). Theoretical considerations underlying Na(+)
uptake mechanisms in freshwater fishes. Comparative Biochemistry and Physiology C
148, 411–418.
Pfaffl, M. W. (2004). Quantification strategies in real-time PCR. In A-Z of quantitative PCR
(Bustin, S. A., ed), pp. 1–23. La Jolla, CA: International University Line.
Richards, J. G., Semple, J. W., Bystriansky, J. S. & Schulte, P. M. (2003). Na+ /K+ -ATPase
𝛼-isoform switching in gills of rainbow trout (Oncorhynchus mykiss) during salinity
transfer. Journal of Experimental Biology 206, 4475–4486.
Sauer, J. & Harrington, J. P. (1988). Hemoglobins of the sockeye salmon, Oncorhynchus nerka.
Comparative Biochemistry and Physiology A 91, 109–114.
Scheffé, H. (1959). The Analysis of Variance. New York, NY: John Wiley & Sons.
Seidelin, M., Madsen, S. S., Blenstrup, H. & Tipsmark, C. K. (2000). Time-course changes in
the expression of Na+ ,K+ -ATPase in gills and pyloric caeca of brown trout (Salmo trutta)
during acclimation to seawater. Physiological and Biochemical Zoology 73, 446–453.
Sigholt, T. & Finstad, B. (1990). Effect of low temperature on seawater tolerance in Atlantic
salmon (Salmo salar) smolts. Aquaculture 84, 167–172.
Singer, T. D., Clements, K. M., Semple, J. W., Schulte, P. M., Bystriansky, J. S., Finstad, B.,
Fleming, I. A. & McKinley, R. S. (2002). Seawater tolerance and gene expression in two
strains of Atlantic salmon smolts. Canadian Journal of Fisheries and Aquatic Sciences
59, 125–135.
Stagg, R. M., Talbot, C., Eddy, F. B. & Williams, M. (1989). Seasonal variations in osmoregu-
latory and respiratory responses to seawater exposure of juvenile Atlantic salmon (Salmo
salar) maintained in fresh water. Aquaculture 82, 219–228.
Stefansson, S. O., Björnsson, B. Th., Ebbesson, L. O. E. & McCormick, S. D. (2008). Smolti-
fication. In Fish Larval Physiology (Finn, R. N. & Kapoor, B. G., eds), pp. 639–681.
Enfield, NH: Science Publisher.
Stefansson, S. O., Haugland, M., Björnsson, B. Th., McCormick, S. D., Holm, M., Ebbesson,
L. O. E., Holst, J. C. & Nilsen, T. O. (2012). Growth, osmoregulation and endocrine
changes in wild Atlantic salmon post-smolts during marine migration. Aquaculture 362,
127–136.

© 2014 The Fisheries Society of the British Isles, Journal of Fish Biology 2014, 85, 1163–1176