ICES mar. Sei. Symp., 198: 626-631. 1994
Survival of eggs and yolk-sac larvae of Baltic cod (Gadus
morhua L.) at low oxygen levels in different salinities
Anders Nissling
Nissling, A. 1994. Survival of eggs and yolk-sac larvae of Baltic cod (Gadus morhua
L.) at low oxygen levels in different salinities. - ICES mar. Sei. Symp., 198: 626-631.
In the Baltic Sea, successful spawning of cod is restricted to the deep basins, the
Bornholm, Gdansk, and Gotland basins, with salinities of 10-17. Deep water is
exchanged mainly during periods of inflow of saline water from the North Sea. These
inflows are irregular and stagnant waters prevail for long periods accompanied by
unfavourable oxygen conditions. Consequently, cod eggs occur regularly at low
oxygen levels. In this study, batches of eggs (2, 4, and 7 days old) and larvae (newly
hatched and 4 days old) were exposed to low oxygen concentrations - 1.0-3.5 mg 0 2/ l at two salinities - 11 and 15 - and compared with survival in controls at 100% oxygen
saturation in short-term experiments. Egg survival decreased with decreasing oxygen
concentration, with no difference in tolerance between egg stages. Egg survival at low
oxygen levels was clearly affected by salinity, with significantly higher survival at 15 ppt
than at 11 ppt. This implies that egg survival at low oxygen levels differs between the
spawning grounds of Baltic cod. It is higher in the Bornholm Basin with higher
salinities, 13-17, than in the Gdansk and Gotland basins, with 10-13. For both larval
stages, a marked decrease in tolerance to low oxygen levels was evident between 3.0
and 2.5 mg 0 2/l irrespective of salinity, suggesting that the limit for survival of Baltic
cod larvae is at about 2.5-3.0 mg 0 2/l.
Anders Nissling: Department o f Systems Ecology, Stockholm University, S-106 91
Stockholm , Sweden.
Introduction
Successful spawning of Baltic cod is restricted to the
Baltic deep basins - the Bornholm, Gdansk, and G ot­
land basins. In the Baltic deep basins there is a halocline
at 50-70 m depth with a denser more saline (10-17 ppt)
deep water which is only partly mixed with the less saline
(6-8 ppt) surface water. The deep water is exchanged
mainly during occasions of inflow of saline water from
the North Sea. These inflows are irregular, however,
and during periods between inflows stagnant conditions
prevail with a decrease in the salinity and oxygen content
of the water. These varying conditions have a profound
impact on the reproductive success of fishes with pelagic
eggs. In the Baltic, cod eggs often occur in layers with
low oxygen content (see Lebedek, 1978; Wieland,
1989), as determined by their buoyancy.
The year-class strength of Baltic cod is thought to be
established during the embryonic period and apparently
determined by the thickness of the spawning layer
(Graum an, 1973), i.e., the layer with 11 ppt salinity or
more (Westin and Nissling, 1991) and sufficient oxygen
content for egg and larval development.
The last two major inflows of saline water to the Baltic
Sea occurred in the winters 1975-1976 and 1976-1977
(Fonselius, 1988) and resulted in favourable spawning
conditions. Consequently, an increase in the Baltic cod
stock followed and high catches were reported from the
late 1970s until the mid-1980s [450 000 t in the early
1980s (Hansson and Rudstam, 1990)]. However, since
stagnant water has prevailed in the deep basins from
the late 1970s, there has been a dramatic decrease in the
Baltic cod stock, with poor catches reported since the
mid-1980s [about 75000 t reported in 1992 (P.-O. Larsson. pers. comm.)].
The metabolic rate of fish embryos is independent of
the water oxygen content until a certain critical level is
reached, below which metabolism becomes dependent
on ambient oxygen concentrations (Rombough, 1986).
Any deviation from normal oxygen uptake will influence
development and survival negatively (Serigstad, 1987).
Metabolic rate in fishes is highly influenced by tem pera­
ture, since temperature controls the rate of different
biochemical reactions, and by salinity, which affects
osmoregulation. Consequently, oxygen requirements of
eggs and larvae are affected by tem perature [see Schur-
ICES m ar. Sei. Symp.. 198(1994)
mann and Steffensen, 1992 (juvenile cod); Alderdice
and Forrester, 1971 (Pacific cod eggs); Almatar, 1984
(herring and plaice larvae)] and possibly by salinity (see
Alderdice and Forrester, 1971; Alm atar, 1984).
The aim of the present investigation was to study the
tolerance of Baltic cod eggs and larvae to low oxygen
concentrations at prevailing salinities of the Baltic deep
basins. In short-term experiments (48 h), eggs (three
different stages) and yolk-sac larvae (two stages) were
incubated at low oxygen levels; 1.0-2.8 (for eggs) and
1.5-3.5 (for larvae) mg 0 2/l at 7°C at 11 and 15 ppt and
compared with survival in controls at 100% oxygen
saturation.
Survival o f eggs and yolk-sac larvae o f Baltic cod
627
1a.
15 ppt
0
C
11 ppt
60
control
ox yg en
(mg/l)
Materials and methods
Spawning cod were caught in gillnets at 50-90 m depth
off Gotland (58°N19°E) in the middle of the Baltic Sea.
Eggs and semen were obtained by stripping. Fertiliz­
ation was carried out in water at 7°C and 17 ppt prepared
from filtered (0.2 /im cartridge filter) sea water (7 ppt)
and synthetic sea salt (hw Marinemix). Eggs at three
different stages of development (2, 4, and 7 days old)
and larvae at two stages [newly hatched and 4 days old
(onset of first feeding)] were incubated. W ater of differ­
ent oxygen concentrations was obtained by bubbling
with nitrogen gas to the nearest 0.1 mg 0 2/l (WTW
Oximeter 196). In all incubations, filtered water was
treated with antibiotics [Mycostatin (2500 IU/1), Strep­
tomycin (0.05 g/1), and Doctacillin (0.2 g/1)]. Treated
water was transferred to the incubation flasks by means
of a tube before the eggs and larvae were added. Prior to
incubation the eggs and larvae were kept in water at 7°C
and 17 ppt salinity.
Batches of eggs and larvae were incubated at five
oxygen concentrations, ranging between 1.0 and 2.8 mg
0 2/l for eggs and 1.5 and 3.5 mg 0 2/l for larvae, at two
different salinities; 11 and 15 at 7°C. As controls, incu­
bations at 100% oxygen saturation (10-11 mg 0 2/l) in
respective salinities were used. Each treatment, contain­
ing 125-150 eggs or 10-15 larvae, was incubated for 48 h
in a 500 ml flask with conical stopper to prevent air
bubbles at the start of incubation. The water was mixed
gently three times a day by rotating the flask to prevent
microstratification of oxygen. Each set of incubation,
i.e., 12 flasks, was performed in replicates with batches
from different females as follows: 6 replicates for 2 and 4
days old eggs; 2 replicates for 7 days old eggs; 5 and 6
replicates for newly hatched and 4 days old larvae,
respectively. The flasks were placed on their side and the
eggs developed at the bottom of 11 ppt, since Baltic cod
eggs are non-buoyant at 11 ppt (see Nissling and Westin,
1991); because of positive buoyancy they developed at
the top at 15 ppt. A t the end of the incubations egg
1b.
100
TT
0
80
control
//
2,4
2,0
ox yg en
1c.
1,7
1,4
11 ppt
1,0
(mg/l)
100
60
1
c ontrol
! /
2 ,4
2,0
ox yg en
1,7
B
15 ppt
0
11 ppt
1.4
(mg/l)
Figure 1. Survival of Baltic cod eggs incubated at low oxygen
levels for 48 h at 7°C with salinities of 11 and 15. Survival in
controls is given as absolute survival and survival at low oxygen
concentrations is given as relative survival, i.e. as percentage of
survival in controls, (a) 2 days old eggs (six replicates), (b) 4
days old eggs (six rcplicates), (c) 7 days old eggs (two repli­
cates). Standard deviation shown at top of bars.
628
ICES mar. Sei. Symp., 198 (1994)
A. Nissling
Table 1. Oxygen concentrations (mg 0 2/l) at the start and end of exposure of Baltic cod eggs (2 and 4 days old) (six replicates) and
larvae (4 days old) (one larval group) to low oxygen levels. Each experiment was carried out at 7°C with a salinity of 11 for 48 h.
Larvae
Eggs
End of incubation
(mg 0 2/l)
Start of incubation
(mg (V I)
10-11 (control)
2.8
2.4
2.0
1.7
1.4
1.0
2 days old
4 days old
9.0
2.6
2.3
2.0
1.6
1.4
9.3 ± 0.4
± 0.6
± 0.1
± 0 .1
± 0.1
± 0.1
±0
-
-
2.3
1.9
1.6
1.4
1.1
survival and larval condition were checked. White eggs
(dimpled) and eggs which had ceased in development
were considered dead and larval condition was judged as
follows: dead, moribund (partly white), alive but in­
active (did not react when touched), or active (swim­
ming). In order to exclude differences in survival due to
differences in egg quality caused by maternal effects and
stage during the spawning period (see Kjesbu, 1989;
Kjesbu et al., 1991; Solemdal et al., 1991), egg survival
was calculated in relation to survival in the controls
(survival in the controls treated as 100% survival) and
given as a percentage of survival in the controls. To
check the consumption of oxygen during the incubation,
oxygen concentrations at the end of incubations at 11 ppt
were measured.
To examine whether transference from a high salinity
(17) to lower salinities (11 and 15) influenced egg sur­
vival at low oxygen concentrations, two egg batches
were fertilized and incubated at respective salinity until
start of exposure, i.e., day 3 (one batch) and day 5 (one
batch), and then incubated at low oxygen concentrations
at 7°C for 48 h as above.
±
±
±
±
±
0.1
0.1
0.1
0
0.1
Start of incubation
(mg 0 2/l)
End of incubation
( m g 0 2/l)
2 days old
10-11 (control)
3.5
3.0
2.5
2.0
1.5
-
9.6
3.4
3.0
2.5
2.0
1.5
—
developmental stages studied, with no apparent differ­
ence in tolerance to low oxygen levels between egg
stages. Com pared with controls (100% oxygen satu­
ration) there was a decrease in egg survival at 2.8 mg O2/
1 and only a few eggs survived at 1.0 mg 0 2/l. However,
egg survival differed between salinities and was signifi­
cantly higher at 15 ppt than at 11 ppt, p < 0.001 (sign test
of paired observations) for both 2 and 4-day-old eggs
(the numbers of incubations of 7-day-old eggs were too
low for a sign test) at low oxygen concentrations, while
no difference in survival occurred between controls at 11
and 15 ppt.
Table 2 gives egg survival at low oxygen concen­
trations at 11 and 15 ppt for two egg batches incubated at
the same salinities from fertilization until the start of
exposure to low oxygen concentrations. Egg survival
was higher at 15 ppt than at 11 ppt for both 3 and 5-dayold eggs. These results suggest that the difference in egg
survival between salinities was not caused by the trans­
ference of eggs from a higher salinity (17 ppt) at the start
of exposure.
In short-term incubations of larvae at low oxygen
concentrations at 11 and 15 ppt, no differences in toler­
ance to low oxygen concentrations were apparent be-
Results
Table 1 gives the oxygen concentrations at the end of
incubations for 2 and 4-day-old eggs and one larval
group. The measurements revealed a small decrease in
oxygen content during the 48 h incubation. The highest
decrease occurred in the controls, while the decrease
was low at low oxygen levels. The decreases in oxygen
content seem to be related to egg survival, i.e., the
higher the survival the higher the oxygen consumption
(cp. with Fig. 1).
Egg survival and larval condition after incubation at
low oxygen concentrations at 11 and 15 ppt are shown in
Figures la - c and 2 a-d, respectively. Egg survival de­
creased with decreasing oxygen concentrations at all
Table 2. Survival of Baltic cod eggs (3 and 5 days old) fertilized
and incubated with salinities of 11 and 15 and exposed to low
oxygen levels for 48 h at 7°C at identical salinities.
5 days old
3 days old
Oxygen (mg 0 2/l)
11 ppt
15 ppt
11 ppt
15 ppt
10-11 (control)
2.4
2.0
1.7
1.4
1.0
96.5
34.5
16.1
3.5
0
0
95.5
67.8
36.1
31.8
73.6
22.5
5.0
4.1
1.1
-
70.4
58.8
20.7
9.4
7.5
-
-
7.2
Survival o f eggs and yolk-sac larvae o f Baltic cod
ICES m ar. Sei. Symp., 198 (1994)
□
629
alive active
E3 alive inactive
co ntrol
3,5
3,0
oxy gen
Fig.
2,5
2,0
H
moribund
I
dead
60-
D
alive active
50-
□
alive inactive
1,5
control
3,5
(mg/l)
3,0
oxy gen
2c.
2,5
2,0
1,5
(mg /l)
Fig. 2d.
100
90
80
70-
4030-
9
moribund
■
dead
20
10-
control
3,5
3,0
oxy ge n
2,5
2,0
0
control
3,5
(mg /l)
3,0
oxy gen
(mg/l)
Figure 2. Condition of Baltic cod larvae incubatcd at low oxygen levels for 48 h at 7°C. (a) Newly hatched larvae incubated with a
salinity of 11 (five replicates), (b) newly hatched larvae incubated with a salinity of 15 (five replicates), (c) 4 days old larvae
incubated with a salinity of 11 (six replicates), (d) 4 days old larvae incubated with a salinity of 15 (six replicates).
tween larval stages or between incubation salinities. For
both newly hatched larvae and larvae at the onset of
feeding the majority of larvae were alive and active at 3.5
mg 0 2/l, with inactive and moribund larvae beginning to
appear at 3.0 mg 0 2/l. For both larval stages there was a
marked decrease in tolerance to low oxygen concen­
trations between 3.0 and 2.5 mg 0 2/l. Incubations at 3.0
mg 0 2/l displayed high survival, whereas incubations at
2.5 mg 0 2/l or lower revealed dead or moribund larvae
with only a few surviving larvae at 2.5 mg 0 2/l.
Discussion
Surveys in the Baltic deep basins have shown that adult
cod may occur in layers with oxygen content as low as
1.4-2.9 mg 0 2/l (see Berner and Schemainda, 1957,
1958; Tiews, 1970) and that cod eggs are regularly found
at low oxygen levels; 1.7-5.4 mg 0 2/l (Lebedek, 1978),
2.6-5.4 mg 0 2/l (Wieland, 1989) with high mortalities
reported during the egg stage period; 79-98.8% (Grauman, 1973), 99.8% (Wieland, 1987). The critical oxygen
630
ICES mar. Sri. Sym p., 198 (1994)
A. Nissling
level for successful development of Baltic cod eggs has
been suggested as being 1.4 mg 0 2/l (G raum an, 1973);
> 2.9 mg 0 2/l (Kosior and Netzel, 1989), but experimen­
tal data are scarce. (Oxygen levels reported in ml 0 2/l
are recalculated to mg 0 2/l).
The present study shows that cod eggs may survive at
low oxygen levels for at least two days, but that survival
is greatly affected at oxygen concentrations of 1.0-2.8
mg 0 2/l. Since cod eggs are regularly found at oxygen
levels of this magnitude, these results suggest that low
oxygen levels in the Baltic deep basins may cause high
egg mortalities. As concluded by Wieland, from investi­
gations on cod egg distribution in the Bornholm Basin,
vertical occurrence of eggs is not only controlled by egg
buoyancy. In addition to water density, the vertical
occurrence of eggs is also influenced by oxygen levels as
a result of high egg mortality at low oxygen concen­
trations. Cod eggs were most abundant at depths of 6075 m at oxygen levels of 2.6-6.9 mg 0 2/l, but eggs also
occurred deeper, at oxygen concentrations of less than
1.4 mg 0 2/l (Wieland. 1987, 1989).
According to several investigations on marine fish
embryos the uptake rate of oxygen increases during
development from fertilization to hatching and absorp­
tion of the yolk sac [Serigstad, 1987 (cod); Davenport and
Lönning, 1980 (cod); Finn etal., 1991 (halibut)], and is
also influenced by larval activity level (Davenport and
Lönning, 1980; Serigstad, 1987), i.e., an increase of
oxygen demand during development. However, in this
study no differences in tolerance to low oxygen levels
among egg stages or between two larval stages were
observed. For both larval stages a marked decrease in
tolerance to low oxygen concentrations was evident
between 3.0 and 2.5 mg 0 2/l, with only a few surviving
larvae at 2.5 mg 0 2/l. Accordingly, this suggests that the
limit for survival of Baltic cod larvae is at about 2.5-3.0
mg 0 2/l.
Incubation of eggs and larvae at two different salini­
ties showed that larval condition was not affected by
salinity at low oxygen levels, whereas egg survival at low
oxygen concentrations was significantly higher at 15 ppt
than at 11 ppt. Hence, the higher mortality of eggs at low
oxygen levels at 11 ppt compared to 15 ppt, whereas no
differences in egg mortality occurred between controls
(100% oxygen saturation), suggests that cod eggs are
more vulnerable to oxygen deficit at extreme salinities.
Further, this investigation indicates that Baltic cod are
less affected by extremes in salinity as larvae than at the
egg stage.
Although a long-term experiment is required to esti­
mate overall egg mortality from fertilization to hatching
at a certain oxygen level, the present investigation
indicates a significant difference in total egg mortality
between 11 and 15 ppt. For instance, if egg mortality at
48 h of exposure at 2.4 mg 0 2/l (egg mortality at the
three egg stages pooled for 11 and 15 ppt) is extrapo­
lated, it results in an overall mortality (egg development
time 13-15 days at 7°C) of about >99.99 and 97% at 11
and 15 ppt, respectively. This indicates a considerable
difference in spawning success with salinity for a given
oxygen concentration.
The inflows of saline oxygen rich water from the
North Sea are irregular and as a consequence conditions
for successful spawning vary greatly between years, but
also between spawning areas. Because of the relative
distance from the North Sea, the most favourable con­
ditions, generally higher oxygen concentrations and
more saline water (13-17 ppt compared with 10-13 ppt
in the Gdansk and Gotland basins), occur in the Born­
holm Basin, which is more regularly influenced by
influxes of water from the North Sea than are the
Gdansk and Gotland basins, where stagnant periods are
more pronounced. Consequently, egg survival is higher
in the Bornholm Basin than in the Gdansk and Gotland
basins, owing to more favourable oxygen conditions but
also, as this study implies, because of higher salinity,
since eggs are less vulnerable to oxygen depletion at 15
ppt than at 11 ppt.
A cknow ledgm ents
The investigation was financially supported by the
Swedish Council for Forestry and Agricultural Research
and the National Swedish Board of Fisheries.
References
Aldcrdice, D. F., and Forrester, C. R. 1971. Effects of salinity,
tem perature, and dissolved oxygen on early development of
Pacific cod (Gadus macrocephalus). Fish. Res. Bd C an., 28:
883-902.
Alm atar, S. M. 1984. Effects of acute changes in tem perature
and salinity on the oxygen uptake of larvae of herring
(Clupea harengus) and plaice (Pleuronectes platessa). Mar.
Biol., 80: 117-124.
Berner, M ., and Schemainda, R. 1957. Ü ber den Einfluss der
hydrographischen Situation insbesondere des Durchlüf­
tungszustandes - auf die vertikale Verteilung und den Fang
der Laichdorschschwärme im Bornholmbecken. Z. Fisch.
N.F., 6: 331-342.
Berner, M., and Schemainda, R. 1958. Ü ber die Abhängigkeit
der Laich dorscherträge im Bornholmbecken von der hydro­
graphischen Situation. Dt. Fischerei Z ., 5(3): 65-70.
D avenport, J., and Lönning, S. 1980. Oxygen uptake in devel­
oping eggs and larvae of the cod, Gadus morhua L. J. Fish
Biol., 16: 249-256.
Finn, R. N., Fyhn, H. 1., and Evjen, M. S. 1991. Respiration
and nitrogen metabolism of Atlantic halibut eggs (H ippoglossus hippoglossus). Mar. Biol., 108: 11-19.
Fonselius, S. 1988. Long-term trends of dissolved oxygen, pH
and alkalinity in the Baltic deep basins. ICES CM 1988/C:23,
p p . 1 -6 .
G raum an, G. B. 1973. Investigations of factors influencing
fluctuations in the abundance of Baltic cod. Rapp. P.-v.
Réun. Cons. int. Explor. M er, 164: 73-76.
ICES mar. Sei. Symp., 198 (1994)
Hansson, S., and Rudstam, L. G. 1990. Eutrophication and
Baltic fish communities. Ambio, 19: 123-125.
Kjesbu, O. S. 1989. The spawning activity of cod, Gadus
morhua L. J. Fish Biol., 34: 195-206.
Kjesbu, O. S., Klungsøyr. J., Kryvi, H.. Witthames. P R., and
G reer Walker, M. 1991. Fecundity, artesia, and egg size of
captive Atlantic cod (Gadus morhua L.) in relation to proxi­
mate body composition. Can. J. Fish, aquat. Sei., 48: 23332343.
Kosior, M.. and Nctzcl. J. 1989. Eastern Baltic cod stocks and
environmental conditions. Rapp. P.-v. Réun. Cons. int.
Explor. M er, 190:159-162.
Lebedek, W. 1978. Occurrence and distribution of sprat and
cod eggs and larvae in the southern Baltic in 1977. ICES CM/
J:15, pp. 1-4.
Nissling, A ., and Westin, L. 1991. Egg buoyancy of Baltic cod
(Gadus m orhua) and its implications for cod stock fluctu­
ations in the Baltic. Mar. Biol., I l l : 33-35.
Rombough, P. J. 1986. Mathematical model for predicting the
dissolved oxygen requirements of steelhead (Salmo gairdneri) embryos and alevins in hatchery incubators. Aquacul­
ture, 59: 119-137.
Schurmann, H ., and Steffensen, J. F. 1992. Lethal oxygen
Survival o f eggs and yolk-sac larvae o f Baltic cod
631
levels at different tem peratures and the preferred tem pera­
ture during hypoxia of the Atlantic cod, Gadus morhua L.
J. Fish. Biol., 41: 927-934.
Serigstad, B. 1987. Oxygen uptake of developing fish eggs and
larvae. Sarsia, 72: 369-371.
Solemdal, P., Kjesbu, O. S., Opstad, I., Skiftesvik, A. B.,
Birkeland, R ., Bratland, P., and Fonn, M. 1991. A compari­
son of egg quality and larval viability between eultured
coastal cod and wild Arcto-Norwegian cod. ICES CM/F:41,
pp. 1-13.
Tiews, K. 1970. Ü ber die Verbreitung des Laichdorschbes­
tandes in der mittleren Ostsee in Abhängigkeit vom Sauer­
stoffgehalt in den Jahren 1962-1970. Arch. Fischereiwiss.,
21:213-221.
Westin, L., and Nissling, A. 1991. Effects of salinity on sperm a­
tozoa motility, percentage of fertilized eggs and egg develop­
ment of Baltic cod (Gadus m orhua), and implications for cod
stock fluctuations in the Baltic. Mar. Biol., 108: 5-9.
Wieland, K. 1987. Distribution and mortality of cod eggs in the
Bornholm Basin (Baltic Sea). ICES CM/G:56, pp. 1-14.
Wieland, K. 1989. Small-scale vertical distribution and m or­
tality of cod and sprat eggs in the Bornholm Basin (Baltic Sea)
during two patch studies in 1988. ICES CM/J:19, pp. 1-15.