migratory Atlantic salmon, brown trout and brook trout in acidic aluminium-rich soft water. Rosse½~nd, B.O. I), Skogheim, O.K. I), Kroglund, F. I) and E. Hoell j.
Mortality and physiological stress of year-classes of landlocked and migratory Atlantic salmon, brown trout and brook trout in acidic aluminium-rich soft water. Rosse½~nd, B.O.I), Skogheim, O.K.I), Kroglund, F.I) and E. Hoell j Directorate for Nature and Management, Fish Research Division, Tungasletta 2, N-7000 Trondheim, Norway. I) Present address: Norsk Biotech A/S, P.O. Box 788, Krossen, N-4301Sandnes, Norway. 2) Norcem A/S, N-3950 Brevik ABSTRACT. Physiological stress, measured as changes in plasma chloride, and mortality were measured on different year-classes of landlocked and migratory Atlantic salmon, two strains of brown trout, and brook trout, in a flow-through system with acidic Al-rich soft water. The oldest year-classes of salmon were smolts. Water from Lake Byglandsfjord (pH = 5.9), was enriched with inorganic AI (as AICI3) ~nd H2SO4 to pH = 5.1, total AI = 225 ug L -I, and labile AI = 135 ug L -~. As a reference, lake water was limed by means of a shellsan~ filter to pH = 6.2, increasing Ca-concentration from 1.0 to 1.5 mg L -~. During the 83 hr experiment, neither mortality nor physiological stress occurred in any species or year-class in the limed water. In the acid water, no mortality occurred on any stage of brown trout or brook trout. Among the migratory and landlocked salmon, however, 5% of the alevins died after 49 and 70 hr, respectively. All smolts of both the landlocked and the migratory salmon died after 83 and 35 hr, respectively, the co~responding loss rate of plasma chloride was -0.76 and -1.26 meq CI hr -~. Brook trout, however, increased plasma ion concentration during the experimental period, and hence showed no stress.
i.
INTRODUCTION
Between species and year-classes of salmonids, the smoltification period of migratory Atlantic salmon (Salmo salar L.) seems to be the most sensitive to acidic stress (Rosseland and Skogheim, 1984). The population of l a n d l o c k e d A t l a n t i c s a l m o n in the acidified Lake Byglandsfjorden, southern Norway, has declined during the last decades. In an experiment at Lake Byglandsfjorden, fingerlings of landlocked Atlantic salmon were found to be more sensitive than parr of brown trout (Sa!mo trutta L.) and brook trout (Salvelinus fontinalis Mitchill) (Skogheim e__[tal., 1984). In this study, different year-classes of : I) Atlantic salmon (landlocked and migratory), 2) two strains of brown trout (one strain being genetically selected towards a higher acid resistance ("Genetic", Gjedrem, 1980), and one strain from Lake Byglandsfjorden), and 3) brook Water, Air, and Soil Pollution 30 (1986) 751-756. © 1986 by D. Reidel Publishing Company.
752
B . O . ROSSELAND ET AL.
trout, were exposed to acidic Al-rich soft water. Limed lake water served as reference. Comparison of smolts of the two salmon strains were given special attention. Mortality and changes in plasma chloride were recorded.
2. MATERIAL AND METHODS The experiment was performed at the Bygland Fiskeanlegg, county of AustAgder, southern Norway, in the period May 8-11, 1984. Inorganic A1 (as AiC13) and H2SO 4 were added to water from Lake Byglandsfjorden (LW, pH = 5.9), by use of a peristaltic pump (Ismatech IP-12) in a continous flowthrough system (acidified LW). The added chemicals were mixed by air flushing for lO min prior to entering the fish tanks. Lake water limed by means of a shellsand filter served as reference water (limed LW). Eyed eggs and 1-yr old smolts of migratory Atlantic salmon of River Imsa strain, were transported from the Research Station for Freshwater Fish, Ims, southwestern Norway, two weeks prior to the experiment. Eggs from the same strain had been fertilized and hatched in Lake Byglandsfjord waters, and were used as alevins in this experiment. The landlocked salmon, the two strains of brown trout, and the brook trout, had been raised from eggs in the Bygland Fiskeanlegg. The smolts of landlocked salmon were 2-yr old. Feeding was stopped two days before the experiment. The fish were placed in a flow through system at 4°C, with black PVC-tanks with a volume of 30 L and retention time of 30 min. A semitransparent black cover prevented visual disturbances. Separate tanks were used for mortality and blood sampling experiments. For mortality studies, I0 fish of each year-class of each species were placed in a single tank; for blood sampling, 20 fish per tank of each species were used. The l-yr old brown trout was too small for blood sampling. The mean length (cm) ±SD, weight (g) +__SD and number of fish (n) in the different groups used for blood sampling were: migratory salmon 13.5 cm ± 2.5, 21.5 g ± 3.7 (33), landlocked salmon 14.6 cm ± 0.8, 25.5 g ± 5.4 (25), and brook trout 12.6 cm ± 6.0, 17.4 g ± 2.8 (31), respectively. Eggs and alevins were kept in floating net-cages made of plastic (mesh size 2 mm). At preset intervals, tanks were checked for dead fish. Blood samples were drawn by heart puncture and analyzed as described by Rosseland and Skogheim (1984). Each fish gave rise to only one sample. Water samples were taken on 16 occasions in the acidified water and on lO occasions in the limed water, pH was measured immediately (Orion mod. 207), and aliquots prepared for A1 analyses. All analytical methods are identical to those described by Skogheim and Rosseland (1984).
3. RESULTS pll in the acidified lake water was fairly constant during the 83 hr experiment, with a wean value of 5.14 ( Table I). Calcium concentration was low, 1.02 mg L -~, with no bicarbonate alkalinity. The concentration
M O R T A L I T Y AND P H Y S I O L O G I C A L STRESS OF YEAR-CLASSES
753
TABLE I. Mean value (M) and standard deviation (SD) for some chemical components in the acidified, Al-enriched lake water (LW), and in the limed lake water, N = number of observations. Dissolved organic matter (DOH) was estimated as absorbance of UV-light (at 254 nm).
CO~ON~NTS
M
ACIDIFIED LW ± SD N
pH 5.14 + Ca mg L -I 1.02~ Bicarbonatealkalinity ueq L -I 0 ± Acid-reactive A1 " 228 ± Non-labile A1 " 93 ! Labile A1 " 134 ± DOM cm -I 0.039 ±
0.03 0.04
16 16
0 5 4 8 0
16 16 16 16 16
M
LIMED LW ~ SD
6.18 ~ 0.05 1.53 ! 0.05 35 iii ! 92 ± 20 ! 0.039 ±
4 5 5 0
N I0 i0 i0 i0 i0 I0 i0
of acid-reactive A1 was 278 ug L -I, with 143 ug L -I in the labile A1 form. Fluoride was 46 u~ L -~ for random samples. in water which passed through the shellsand filter, pH increased to a mean of 6.18~ Ca to 1.53 mg L -I, and the bicarbonate alkalinity was 0.035 meq L -~ (Tabl~ I). The concentration of acid-reactive A1 was Iii ug L -I and 20 ug L -l was in the labile A1 form. The concentration of dissolved organic matter (DOM) was low and the same in the two w a t e r ~ Water colour estimated against a Pt-standard was lower than 5 mg Pt L -~ in all samples. The estimated equilibrium concentrations of labile A1 with respect to a solid phase with similar solubility as synthetic gibbsite, indicate that in the a c i d i f i e d lake water there was a slight undersaturation (Table II). In the limed lake water, however, a slight oversaturation occurred. In the acidified water, the AIF z+ was the dominating Al-species, whereas aluminat~ ion (AI(OH)4-), A1 hydroxide (AL(OH)2+) and A1 trihydroxide (AI(OH)3U) were present in the limed water (Table II). During the 83 hr experiment, no mortality occurred in any species or year-class in the limed water. In the acidified lake water, no mortality occurred on any stage or strain of brown trout or brook trout. However, both the migratory and landlocked salmon experienced a 5% mortality of alevins after 49 and 70 hr, respectively. One parr (10%) of the one-year old landlocked salmon died after 83 hr. Both the migratory and landlocked salmon smolts experienced 100% mortality after 35 hr (LT50A~ 25 hr) and 83 hr (LT 50 = 70 hr), respectively. the start of the experiment, plasma chloride concentration differed between groups; m~an concentration + S D in landlocked Atlantic salmon was 107 + 1 meq L -~ (n = 3), versus--99 + 8 meq L -i (n = 8) for migratory s a l m o n and 96 ± 9 meq L(n =--4) for brook trout, respectively (Figure IA). Both the migratory and landlocked salmon lost plasma chloride at a rapid rate after exposure to the acidified lake water, A C1 = -1.22 and 0.76 meq hr -~, respectively. Brook trout, however, increased plasma
754
B.O. ROSSELAND ET AL.
TABLE ii Analyzed labile AI and calculated Al-speciation i~ acidified lake water (LW) and limed LW. All concentrations in ug A1 L-% ACIDIFIED LW Analyzed labile AI
LIMED LW
134
20
22 28 29 0 48 6 I
0 0 4 7 0 0 0
Inorganic monomeric AI AI(OH) 3
134 0
ii 9
Inorganic AI
134
20
Equilibrium concentration of inorganic monomeric AI
161
11
Distribution of AI species AL ~+ AI(OH) 2+ AI(OH)2+ AI(QH)~AIF z+ ~ AIFo+ AIS04+
A
140
B
140
120
120
"•
c3) 100
lOO
=o 80
80 (6)
eD
--
a.
60
60
40
40 I
0
l
!
20
i
i
40
I
!
i
60
i
t
0
20
i
i
40
i
i
l
60
HOURS
Fig. I. Plasma chloride of migratory and landlocked Atlantic salmon and brook trout in A) acidified lake water and B) limed lake water. Mean (n=3 if not otherwise stated) and SD are given. Atlantic salmon: migratory ( ~ - - - n ), landlocked ( ~ ) . Brook trout (A- -, -A).
MORTALITY AND PHYSIOLOGICALSTRESSOFYEAR-CLASSES
755
chloride concentration during the firs~ 37 hr, A C I = 0.32 meq hr -I and stabilized around a mean of 116 meq L -~ (Figure IA). In the limed water, the levels of plasma chloride in all groups increased during the experiment (Figure IB).
4. DISCUSSION As to the equilibrium of labile AI, the acidified water was slightly undersaturated with respect to a solid phase with similar solubility as synthetic gibbsite, in contrast to the slight oversaturation in the limed lake water. In laboratory experiments, Baker and Schofield (1982) observed that oversaturated A1 solutions were particularly toxic to salmonids. In connection with liming, however, such oversaturated solutions seems not to cause any stress (Rosseland e__ttal., 1986). No mortality occurred among the two strains of brown trout or the brook trout, whereas all stages (except eyed eggs) from alevins to parr and smolts of the two Atlantic salmon strains experienced different degrees of mortality. This species difference is in accordance with other studies (Grande et al., 1978; Rosseland and Skogheim, 1984; Skogheim e t al., 1984; Fivelstad and Leivestad, 1984). The fact that 100% mortality only occurred in smolts supports the observation that salmon, either migratory or landlocked, are most sensitive during the smoltification period (Rosseland and Skogheim, 1982, 1984). The loss of plasma chloride of smolts in this experiment was more rapid than at comparable pH and labile Al-concentrations described by Rosseland and Skogheim (1984). The low~r Ca- concentration in the experiment reported here (I.0 vs 2.0 mg Ca L-~), is likely to be the explanation for this. The species difference in sensitivity between brook trout and Atlantic salmon is clearly illustrated by the salmon loosing chloride rapidly whereas the brook trout gained chloride. Handling and confinement have been shown to induce loss of plasma chloride in salmonids (Wedemeyer, 1972; Redding and Schreck, 1983). The fact that both salmon and brook trout plasma chloride concentration increased in the limed lake water, however, indicates that confinement in the experimental fish tanks are not likely to have caused any osmotic stress. The migratory salmon had a subnormal level of plasma chloride at the experimental start, maybe due to stress caused by handling and transportation (Soivio and Virtanen, 1984). This might explain the more rapid ion loss and mortality compared to the landlocked salmon. Genetics, however, have been shown to play some role in the tolerance to acid waters in strains of brown trout (Gjedrem, 1980; McWilliams, 1982) and brook trout (Robinson e__qtal.~ 1976). Acclimation, caused by raising of the landlocked salmon in the moderate acidified Lake Byglandsfjord, might also be of importance, as McWilliams (1980a,b), Schofield et al. (1981) and Fivelstad and Leivestad (1984) all found acclimation to be relevant within certain limits. Which of these factors, or combination were the cause of the observed difference in tolerance between landlocked and migratory salmon cannot be determined from this study.
756
B. O. ROSSELAND ET AL.
REFERENCES Baker, J.P. and Schofield, C.L.,: 1982, Water~ Air~ Soil Pollut. 19, 289. -Fivelstad, S. and Leivestad H,: 1984, R ep. Inst. Freshw. Res. Drottningholm 61, 69. -Gjedrem, T.: 1980, 'Genetic variation in acid tolerance in brown trout', in: Drabl~s, D. and Tollan A, (eds.), Ecolo~ical impact of acid precipitation, SNSF-project, pp. 308. -Grande, M., bluniz, I.P., and Andersen S,: 1978, Verh.Int.Verein. Limnol. 20, 2076. -McWilliams, P.G.: 1980a, J..~.Exp. Biol. 88, 259. -McWilliams. P.G.: 1980b, J_-..Exp, Biol. 88, 267. - McWilliams, P.G.: 1982, Comp. Biochem. Physiol. 72a, 515. -Redding, J.M. and Schreck, C.B.,: 1983, Trans. Am. Fish. Soc. 112, 800. - Robinson, G.D., Dunson, W.A., Wright, J.E., and Mamolito, G.E.,: 1976, J. Fish. Biol. 8, 5. - Rosseland, B.O. and Skogheim, O.K.: 1982, 'Physiological stress and mortality of Atlantic salmon, Salmo salar L. in acid water with high levels of aluminium', Inter. Council for Exploration of the Sea C.M. 1982/M: 29, 15 p. -Rosseland, B.O. and Skogheim, O.K.: 1984, Rep. Inst. Freshw. Res. Drottningholm 61, 186. - Rosseland, B.O., Skogheim, O.K., Abrahamsen, H. and Matzow, D.: 1986, 'Limestone slurry reduces physiological stress and increases survival of Atlantic salmon (Salmo salar L.) in an acidic Norwegian river', Can. J. Fish Aquat. Sci. __43, in press. -Schofield, C.L., Webster, D.A., Guthrie, G.A. and Flick, W.A.: 1981, 'Acclimation of brook trout to acidified waters', in: Hanagement of acidified waters. Contr. C-164480, N.Y. State Dept. Environm. Cons. and Cornell Univ., N.Y. -Skogheim, O.K. and Rosseland, B.O.: 1984, Rep. Inst. Freshw. Res. Drottninghglm 61, 17. -Skogheim, O.K., Rosseland, B.O., Hafsund, F., Kroglund, F. and Hagenlund, G.: 1984, 'Eksponering av bleke, aure og bekker~ye i Surt vann' (Exposure of landlocked salmon, brown trout and brook trout in acid water), Rapport fra Fiskeforskningen, 2/84, 14p. (In Norwegian with English summary). - Soivio, A. and Virtanen, E.: 1984, 'Physiological effects of stocking stress on Baltic salmon (Salmo salar)', EIFAC Technical Pape r No 42, Suppl. Vo!. I, 217. - Wedemeyer, G.: 1972, J. Fish Res. Board Can_. 29, 1780. -
