Skip to main content

Selective breeding for stress tolerance in aquacultured fish


The overall aim of this study is to assess the feasibility of increasing the tolerance of finfish to stress selective breeding. Because ideal husbandry practices are of necessity compromised by the economic realities of large-scale fish production, stress is an unavoidable component of the finfish aquaculture environment. Although the physiological response of fish to isolated stressors is essentially adaptive, fish under intensive culture conditions are exposed to repeated acute stress and in some instances to chronic stress. In these cases the response becomes maladaptive resulting in adverse effects in growth, flesh quality, reproductive function and immunocompetence (Pickering, 1993; Campbell et al., 1994; Pickering and Pottinger, 1989; Sunyer et al., 1995). Therefore, an enhanced tolerance of stressful procedures is likely to improve food conversion and growth, reduce disease incidences and improve broodstock performance. Thus, an investigation on selective breeding of fish for stress tolerance has been undertaken in the gilthead sea bream Sparus aurata and rainbow trout Oncorhynchus mykiss. The selection procedure is based on traits, which are strongly linked with responsiveness to stress: Plasma cortisol and lysozyme, as previous work from the laboratories participating in this work estimated the potential heritability of these traits (Fevolden and Roed, 1993; Fevolden et al., 1991, 1994; Roed et al., 1993a,b; Pottinger and Pickering, 1997).

Sea bream and trout were selected for high or low stress response, measured by plasma cortisol levels, and tested after repeated confinement stress. The consistency of the response and the correlation with other indicators was analysed. After this, crosses between either high cortisol (HR) male and female responders and low cortisol responders (LR) were made. Reproductive parameters were measured and correlated with cortisol. The results show that selected fish consistently show the corresponding high or low response and that growth rate is higher in HR. Reproductive parameters show slight differences in trout between HR and LR whereas in sea bream, HR show lower egg quantity but better quality. F1 generation show the same pattern of cortisol responsibility as the respective progenitors.
The results of the selection procedure in terms of the cortisol response has resulted in a consistency, as the fish selected as either high or low response continued showing high and low levels respectively in sea bream and trout. Similarly, in both species a positive correlation was found between growth and high cortisol response. Regarding the differences between species, the results are the following:

Sea Bream:

The results obtained up to this moment regarding reproductive parameters show the following trends: Only 65% of the pairs selected for HR or LR spawned whereas 100% of unselected pairs spawned. No correlation was found between the number of spawns and post-stress plasma cortisol in females. There were no significant differences in the number of eggs per kilogram of female between HR and LR. However, both groups had significantly more eggs per kilogram of female per day than the US group. Egg quality in terms of viability and hatching rates was significantly better for HR than the other groups, whereas lipid droplet diameter was lowest in US group. No significant differences were found for newly hatched larvae length between groups.

The evaluation of the impact of a stressor on the reproductive system of selected fish showed that confinement caused a clear decrease of 17B-estradiol but no differences between HR and LR. No significant differences were found in GSI, number of eggs per female, individual egg weight, sperm counts in males or egg mortality at hatching and eyeing. However, egg volume was lower in LR and sperm count was higher either HR or LR than in unselected fish (US).

The first examination of the stress-responsiveness of the F1 strongly indicates that the fish display the same traits for HR or LR as the parent generation. Up to three assessments on the progeny show high significant differences in post-stress cortisol levels in HR than in LR.

Different results have been found as regards the metabolic response of the progeny of HR or LR fish under chronic stress. Unstressed fish show better growth rate and feed efficiency in LR group, whereas in fish subjected to high density, growth rate and feed efficiency was significantly lower in HR and condition factor was lower for the HR crowded fish. Liver Alanine Aminotransferase showed lower activity in LR both in crowded and uncrowded fish and Piruvate kinase showed higher activity in HR also in both groups. Phosphofructokinase increased in the HR crowded group.

Regarding heretabilities of selected traits as cortisol and lysozyme, the results showed moderate to high heretability estimates, whereas the common full-sib effects due to other than additive effects was negligible.

In conclusion, these preliminary results show that in sea bream and trout, the selection for cortisol appears to be consistent and heritable. Nevertheless, it is not clear whether selection for low cortisol response is always recommendable. Further analysis will be carried out on the consistency of stress/cortisol response on this and next generation to see whether the selection for this characteristics may help to decrease husbandry problems and improve the profitability of aquacultured fish.


Campbell , P.M., Pottinger, T.G. and Sumpter, J.P. (1994). Preliminary evidence that chronic confinement stress reduces the quality of gametes produced by brown and rainbow trout. Aquaculture 120, 151-169.

Fevolden, S.E., Refstie, T. and Roed, K.H. (1991). Selection for high and low cortisol stress response in Atlantic salmon (Salmo salar) and rainbow trout (Oncorhynchus mykiss). Aquaculture 95, 53-65

Fevolden, S.E., Roed, K.H.and Gjerde, B. (1994). Genetic components of post-stress cortisol and lysozyme activity in Atlantic salmon; correlations to disease resistance. Fish Shellfish Immunol. 4, 507-519.

Fevolden, S.E. and Roed, K.H. (1993). Cortisol and immune characteristics in rainbow trout (Oncorhynchus mykiss) selected for high or low tolerance to stress. J. Fish Biol. 43: 919-930.

Roed, K.H., Larsen, H.J., Linder, R.D. and Refstie, T. (1993)a. Genetic variation in lysozyme activity in rainbow trout (Oncorhynchus mykiss). Aquaculture 109, 237-244

Roed, K.H., Fjaelstad, K.T. and Stromsheim, A. (1993)b. Genetic variation in lysozyme activity and spontaneous hemolytic activity in Atlantic salmon (Salmo salar). Aquaculture 114, 19-31

Pickering, A.D. (1993). Growth and stress in fish production. Aquaculture 111: 51-63.

Pickering, A.D. and Pottinger,T.G. (1989). Stress responses and disease resistance in salmonid fish: effects of chronic elevation of plasma cortisol. Fish Physiol. biochem. 7, 253-258.

Pottinger, T.G. and Pickering, A.D. (1997). Genetic basis to the stress response: selective breeding for stress. In: "Fish stress and health in aquaculture" (Iwama, Pickering, Sumpter and Schreck, eds.) pp: 171-193. Cambridge Univ. Press. Cambridge. UK

Sunyer,J.O., Gomez,E. Navarro,V. Quesada,J. and Tort,L. (1995). Depression of humoral components of the immune system and physiological responses in gilthead sea bream Sparus aurata after daily acute stress. Can. J.Fish. Aquat. Sci. 52: 2339-2346.
Material and Methods

Sea bream
Broodstock of sea bream of 900g (147 fish) from both Mediterranean and Atlantic populations were subjected to 3h confinement in their own tanks by restricting the living space with net walls. In this way, space restriction is reached, but avoiding loss of water quality, as water flow and renovation is always maintained After confinement, fish were individually netted, anaesthetised, weighed and blood was sampled through caudal puncture. This procedure was repeated every month during 5 consecutive months (June to October). Plasma and serum was obtained after centrifugation and concentrations of cortisol, glucose, lactate, osmolality, and agglutination activity were determined. Once the fish samples were analysed for cortisol response to the stress, two males and one female of fish of high cortisol response and the same for low cortisol response were placed in individual tanks. The fish were classified as high cortisol responders (HCR) or low cortisol responders (LCR) depending on the consistency of the response to the stress along the five months. The offspring of these families was analysed in terms of quantity and quality of the eggs. The variables measured were: Number of eggs per female, spawning viability, percentage of hatched eggs and larval size after hatching.

Two year-old mixed sex rainbow trout were individually identified and subjected to 3 hours confinement (10 fish per 50 l tank) at monthly intervals. After confinement fish were transferred to anaesthesia and 1 ml of blood was removed. Plasma cortisol, glucose and lysozyme levels were determined. To perform the crosses, 15 females from the HR and LR groups were made with the corresponding 15 males. In addition, 15 males and females randomly chosen from the US group (Unselected fish) were mated. Total egg volume and egg size, egg mortality and sperm count were determined

Funding Scheme

CSC - Cost-sharing contracts


The Ferry House Far Sawrey
United Kingdom

Participants (5)

Apartado De Correos, 56 Muelle De Taliarte S/n
Ullevaalsveien 72
0033 Oslo
Tehag Fish Centre
Vorosmarty Street 68
2441 Szazhalombatta
201 B,breivika
9037 Tromsoe
Universidad de Granada
Campus Fuente Nueva S/n
18071 Granada