AquaBreeding Project title: “Towards enhanced and sustainable use of genetics and breeding in the European aquaculture industry”

Дата канвертавання24.04.2016
Памер1.25 Mb.
1   2   3   4   5   6   7   8   9   ...   16

Reviews on

Breeding and Reproduction of

European aquaculture species

Review on Breeding and Reproduction of European aquaculture species

Rainbow trout (Oncorhynchus mykiss)

Marc Vandeputte (INRA, France), Arne Storset (Aquagen AS, Norway), Antti Kause (MTT Agrifood Research, Finland), Mark Henryon (Aarhus University, Denmark)

January 2008

1. General information on production and breeding

History of the domestication

Rainbow trout originate from North America and was one of the first species of fish to be domesticated. The history of trout domestication has been extensively reviewed by Gall and Crandell [1]. The first rainbow trout hatchery was established in 1879 on the McCloud River in California. It produced over 2 million eggs until its closure in 1888. The initially spring-spawning rainbow trout was selected in order to produce winter, fall and even summer-spawning fish. The species spread throughout the world during the 19th and 20th centuries. European trout farming began in Denmark in the 1890s. Rainbow trout farming is now much more important in the rest of the world than in its native North America.

Production (Figure 1)

Worldwide production of rainbow trout was 537,000 tonnes in 2005 [2]. Europe is the largest producer with 273,000 t. South America produces 132,000 t, Asia 97.000 t, and North America 32,000 t. The largest producers are Chile (118,000 t), Norway (59,000 t), Turkey (50,000 t), and Denmark, Iran, France and Italy (30-40,000 t each). Freshwater production is most common in Europe; with production of 172,000 t portion-sized trout per year [3]. Saltwater production is mainly done in Norway, Finland and Denmark. These countries produce approx. 77,000 t of 2-4 kg fish per year.

Figure 1: Production countries in the Euro-Mediterranean region.

Biological features of interest for breeding practices

  • Main commercial data on weight and age at harvest and slaughtering performances

There are two types of trout production. The first is freshwater production in raceways of portion-sized fish for human consumption (i.e., fish with a live body weight of 200-400 g). The second is production of 2-4 kg fish for meat and caviar production, in sea cages or in freshwater raceways.

Portion-sized trout reach commercial size after 6-15 months post-fertilisation. Trout require 15-36 months to reach 2.5 kg. The large variation in growth duration is due to the genetics of the fish and numerous environmental factors, one of the most important being water temperature.

Sexual maturation of the trout reduces product quality. Male maturation causes the greatest problems. Males can mature prior to reaching portion-size. Female maturation, on the other hand, only causes problems for the production of 2-4 kg fish. For this reason, many producers use monosex female populations.

  • Generation interval for males and females

Rainbow trout males mature between 1 and 3 years, depending on strain and temperature regime. The first maturation of females occurs between 2 and 4 years of age.
Main Diseases

Bacterial, viral, fungal, and parasitic diseases are a major constraint for trout production.

Bacterial diseases: furunculosis (causative agent: Aeromonas salmonicida), enteric redmouth disease (Yersinia ruckeri), Bacterial kidney disease (Renibacterium salmoninarum), and rainbow trout fry syndrome (Flavobacterium psychrophilum). Vibriosis (Vibrio anguillarum) can also affect rainbow trout, but in sea water only. Vaccines exist only for furunculosis, enteric redmouth disease and vibriosis.

Viral diseases: The major viral diseases affecting rainbow trout are VHS (Viral Haemorrhagic Septicaemia), IHN (Infectious Haematopoietic Necrosis) and IPN (Infectious Pancreatic Necrosis). Outbreaks of the first two diseases require declaration to the authorities, according to the EU Council Directive 91/67/EEC defining Approved Zones, free of such diseases.

Fungal: The main fungal infections are caused by Saproleignia spp. They are prevalent during the reproductive season, when males are particularly sensitive.

Parasitic: The main parasitic disease is caused by Gyrodactylus derjavini.

The so-called “strawberry disease” is quite frequent in some farms. As yet, its aetiology is unknown, but it may be relieved by antibiotherapy [4].

2. Genetic variability of the species

Wild genetic resources available

The natural distribution area of the rainbow trout is along northern coasts of the Pacific Ocean, from Kamchatka to California [5]. There is a major differentiation between coastal, inland and southern Californian populations [5,6]. In Europe, only a few feral populations of rainbow trout exist in Denmark and Norway. So, it is unlikely that these provide an important gene pool suitable for commercial production.

Differences between wild and/or domesticated populations

Many of the cultured strains of trout derive from the initial stock of the McCloud river hatchery in the 1880s [5,7]. Hatchery strains often show low levels of heterozygosity, although this is not always the case. Some French strains of rainbow trout exhibit higher variability than many wild US strains [8].

Studies made essentially in Europe comparing rainbow trout strains for production traits revealed large variation for survival, growth rate, body composition, processing yields, and age at maturity [9-15]. In a study where a wild strain was compared with a number of hatchery strains, growth performance of the wild strain was only half that of an average hatchery strain [10]. Such differences illustrate the high divergence between wild and domesticated strains. Large behavioural differences also exist between wild and domesticated strains. Domesticated strains show much lower anti-predator and higher agonistic behaviour [16-18]. Therefore, there is much genetic gain to be made by selecting the best domesticated strains as genetic stock for commercial production, but it is unlikely to find any advantage in using the wild or feral strains.
Interaction between wild and domesticated stocks

The interaction between wild and domesticated stocks of rainbow trout is not an issue in Europe, as no wild stocks exist. The effect of feral trout on local fish populations does not seem to be a serious threat [19], although any introduction should be viewed with caution.

Inbreeding effects

Inbreeding effect on growth is -0.2%to-6% per 10% increase of inbreeding [20-22]. The effect of 0.25-0.50 inbreeding on survival is 10.0% for eyed-eggs, 5.3% for alevins and 11.1% for fry [21]. Inbreeding also causes delayed spawning and reduced egg number [22].

3. Reproduction

Fecundity and main reproductive features

Natural populations of rainbow trout spawn in spring, but strains have now been selected to spawn at all times of the year [1]. Female trout have a fecundity of approx. 2,000 eggs/kg body weight. They will ovulate, but not spawn, in culture conditions. Only artificial fertilization may therefore be performed, with hand stripping of milt and ovulated eggs. The spawning times of the females can be synchronized using hormonal injection (GnRH analogues, Ovaplant TM Gonazon TM – [23]). Photoperiod manipulation allows an almost year-round production of eggs [24]. However, photoperiod manipulation is less feasible in sea-cage production.


Sperm cryopreservation protocols have been published and are effective, although very high thawed sperm concentrations are needed to achieve good fertilization rates. There is high individual variability between in their capacity to endure cryopreservation [25-28]. This allows the use of cryopreservation for conservation of lines, but prevents its use as a routine tool in breeding programmes [25]. Methods for large scale application exist, but are not yet published (see Blastomere cryopreservation is also feasible [25]. Recently, the possibility to obtain rainbow trout eggs by injection of spermatogonia into triploid sterile masou salmon provided a means to cryopreserve and recover whole genomes in rainbow trout [29]. Collections of frozen semen exist in research institutions and breeding companies, and the French National Cryobank is about to add rainbow trout lines to its collections.

Genetic and environmental sex determination, sexual dimorphism

Sex determination is simple in rainbow trout, with an XX/XY (mammalian style) chromosomal sex determining mechanism. Production of monosex female populations by mating of normal females with “neomales” obtained by hormonal phenotypic sex reversal of genetic females is well established [30,31]. Use of gynogenetic offspring to quickly obtain the first monosex population is also possible. Sex dimorphism for growth is in favour of males (+10%), but the advantage of later maturation of females has resulted in monosex progeny being widely used in commercial production (>90% in France) [32].

4. Selection

Genetic variability per trait

Rainbow trout express genetic variation for most, if not all, commercially-important traits [9,12,32-63]. Many traits are moderate-to-highly heritable, although there are exceptions. Heritability estimates for growth rate is on average 0.35 (0.06-0.53). Carcass and fillet yields have heritabilities from 0.29 to 0.50. The heritability of fillet fat content is around 0.25 (0.03-0.47), while the heritability of fillet protein is lower (0.08 on average, 0.03-0.19). The heritability of fillet colour is highly variable, and often low (0.06-0.27), but colour scores recorded using image analysis may be high in some cases (0.36-0.75). Among reproductive traits, spawning date is highly heritable (0.47-0.86), while age at first maturation is only moderately heritable (0.12-0.35). Both relative fertility (number of eggs/kg BW - h²=0.35) and egg size (h²=0.30-0.60) are heritable, but are of little practical interest. The heritability of condition coefficient is variable but often high (mean 0.45, 0.03-0.71). As for external appearance, both number of black spots (h²=0.45) and skin colour (h²=0.29) can be selected for.

Overall survival is weakly heritable (≈ 0.15), while resistance to specific diseases (IHN, VHS, IPN, ERM, RTFS) often display heritabilities between 0.30 and 0.60. Heritability for skeletal deformations varies with incidence. Heritability tends to be moderate only when deformations are frequent, whereas heritability is close-to-zero when deformations are few.

Individual daily feed intake of rainbow trout shows extensive day-to-day variation. However, given that enough repeated records are measured from each fish, heritability of average feed intake can be moderate. As daily feed intake and daily weight gain are very highly correlated, this results in modest heritability for feed efficiency.

These estimates of heritability clearly indicate that selective breeding programs for rainbow trout will be successful, resulting in much genetic gain.
Genetic correlations and undesirable side effects of selection

Growth is favourably correlated with feed efficiency [54,62]. Thus, selection for rapid growth leads to increased feed efficiency. The relation between body weight and tissues lipid content varies from positive to negative, but in most cases tends to be positive (reviewed by [64]). Rapid growth is thus generally related to high lipid percent.

Growth has an unfavourable genetic correlation with condition coefficient (0.29-0.8 with weight, 0-0.4 with length [40,41,56]), and thus selection for growth will result in more rotund fish. The genetic correlation between successive weightings is usually high, but the correlation between juvenile weight and slaughter weight may be very low [36], making early selection for late growth potentially inefficient. The genetic correlation between early and late muscle lipid content is moderate (0.33-0.45 [43]). Age at maturity is negatively correlated with growth, selection for growth thus promotes early maturity [37]. In contrast, no correlation exists between spawning date and growth [35,37]. Genetic correlations between resistance to ERM, RTFS and VHS are low (-0.11A<0.15 [54]), and a negative (-0.33 to –0.14) unfavourable genetic correlation exists between VHS resistance and body weight [54].
These findings suggest that selection for each trait will generally result in a favourable response in many other traits. However, there is need for caution. Genetic correlations involving growth rate are the only reliable correlations we have at present. Therefore, reliable estimates of genetic correlations are needed for other commercially-important traits.
G*E interactions

There are few data for GxE interactions in rainbow trout. However, the studies that have been carried out suggest that they are small. When full-sib families were evaluated in three Scandinavian production systems (fresh water, brackish water and salt water), a genetic correlation of 0.58, indicating G x E interaction, was found between slaughter weights in freshwater and brackish water [65]. The other correlations were higher (0.86 & 0.72).

Genetic correlation between fresh and brackish water production environments for body shape (0.90) and skin spots (0.92 ) is very high, and lower (0.78) for skin colour[66]. For age at maturity, close to unity (0.96) genetic correlation between fresh and brackish water has been found [42]). Across low and high energy diets, the genetic correlations again are high for growth, proximate body composition and fillet colour. For skeletal deformations, genetic correlations between fresh and brackish waster tend to be high and positive, but can be as low as 0.49 and not significantly different from zero [60].
A small but significant interaction between strains and environment was also found for growth and age at maturity in four Swedish commercial farms [14]. In another experiment comparing sea-water farms in Norway, a low but significant family x farm interaction effect was found on growth [67].
These studies suggest that the relative performance of an improved strain of rainbow trout in one environment will not change drastically when they are reared at different commercial farms. However, this premise needs to be validated to justify the development of a single trout strain.
Genetic responses, progress and control lines

There is growing empirical evidence that selective breeding results in genetic gain. Early selection trials reported genetic gain for growth and spawning date, although these trials were carried out without control lines and were therefore potentially influenced by environment effects [68,69]. Kincaid obtained +22% per generation in three generations of selection for body weight at 5 months of age [70]. The Norwegian National breeding programme produced 4.3% increase in growth rate per generation [71], while the Finnish National breeding programme produced 5 to 12% [72]. Genetic gain has also been demonstrated for spawning date [73]. So, there is little doubt that selective breeding is a worthwhile method of improving commercial trout production.

Dominance and intraspecific crossing

In his review of 1992, Gjedrem highlighted that heterosis has been observed in rainbow trout [74]. However, for most commercially-important traits, heterotic effects were small. So, heterosis and crossbreeding are unlikely to play major roles in trout breeding schemes.

5. Polyploidisation and monosexing and hybrids

Triploid induction and performances

Triploid trout can be obtained by heat shock or pressure shock following fertilization [75,76]. Triploids are sterile; females develop no gonads while only male develop small testes [77]. Unlike diploids, triploid fish do not mature sexually, and there is no decrease in flesh quality [78]. However, the growth and survival of triploids is slightly lower than diploid fish [79]. The growth rate of triploid families is strongly correlated with that of diploids from the same families, enabling genetic gain obtained in diploid fish to also be realised in triploid fish. [80].

Tetraploid induction and performances

Tetraploids can be induced by heat or pressure shock [76], but have low growth (50% that of diploids) and survival (30% that of diploids) [81]. Tetraploid males are fertile, and mating them with diploid females gives 100% triploid offspring. However, the fertilisation rate is very low, presumably because the large size of the (diploid) spermatozoa of the tetraploids does not allow them to pass easily through the egg micropyle [81]. This characteristic varies between individual males, and could thus be improved through selective breeding.

Gynogenesis, androgenesis and mitotic clone performances

Both meiotic and mitotic gynogenesis, as well as androgenesis, have been performed in rainbow trout [76]. These techniques have led to the production of clonal lines. Five androgenetic clonal lines are available at Washington University [82], and 15-20 gynogenetic clonal lines at INRA Jouy en Josas [Quillet, pers. comm.]. Clonal lines exhibit a large genetic variability for many kinds of traits [82,83].

Interspecific hybridisation

Many interspecific hybrids, both diploids and triploids, have been tested in rainbow trout [84,85]. However, none have shown sufficient performance to be used in practice.

6. Genomics

The genomic work conducted in rainbow trout has been reviewed in [86]


Tools to evaluate population genetic variability

Many different types of genetic markers (allozymes, mtDNA, microsatellites, etc) have been used to evaluate population variability.

Genetic markers for genealogical traceability

Rainbow trout was one of the first fish species where genealogical traceability using microsatellite markers was tested on a relatively large scale [87,88]. Several multiplexes are available [89], and several labs perform assignments on a commercial basis.

QTL and Marker Assisted Selection

Four genetic maps have been produced [90-93]. QTLs have been identified for upper thermal tolerance [94-96], resistance to IPN [97], to IHN [98] and to the parasite Ceratomyxa shasta [99], post-stress cortisol levels [100], growth [101-103] and development rate [104]. However, none of these QTLs is currently being used in a marker assisted selection programme.


Rainbow trout is the second species of fish in which successful transgenesis has been reported [105]. Phenotypic results were obtained on growth [106-108], sterilization [109] and various metabolic pathways [110,111] in these fish. The main focus of this work has been on growth. However, it appears that even though a spectacular weight gain is obtained when transgenesis is practiced on wild rainbow trout, no response was seen when it was used on domesticated strains [108]. No commercial applications have been made of transgenic trout.

Review on Breeding and Reproduction of European aquaculture species

Atlantic salmon (Salmo salar)

Ashie Norris (Marine Harvest Ireland, Ireland)

January 2008

1. General information on production and breeding

History of the domestication

The two main Norwegian farmed strains, Aquagen and Mowi, which also constitute a significant proportion of all farmed salmon, were founded from wild parents between late 1960s and early 1970s. Mowi strain was founded from four wild Norwegian rivers. Aquagen was founded from up to 40 river populations.

The first selective breeding program of Atlantic salmon in Norway was a family based program initiated at the beginning of the 1970´s [1, 2]. Because farmed Atlantic salmon in Norway has a 4-year generation interval, four genetically distinct populations were established to provide the salmon farmers with genetically improved smolts each year.. In addition, several other breeding programs based on mass selection were started in the 1970-ies and early 1980-ies with sub-populations of similar origins.

The worldwide production of Atlantic salmon was 1220.000 tonnes in 2005 [3], among which: 750.000 in Europe, 375.000 in South America and 95.000 in North America. The major producers are Norway (580.000 t), Chile (375.000 t), UK (Scotland) (130.000 t), Canada (85,000) and US, Ireland and Faeroe Islands (10-30.000 t each).

Biological features of interest for breeding practices

  • Main commercial data on weight and age at harvest and slaughtering performances

Atlantic salmon is anadromeous meaning they spawn in fresh water and migrate to sea after 1 to 7 years depending on the water temperature in the river. After smoltification the smolts which are 30 -50 g migrate to the sea for grow out. They reach sexual maturity after 1 (grilse) to 3 years at sea and then return to freshwater to spawn. A special feature of Atlantic salmon is a low frequency of precocious males maturing in the freshwater about 20 - 40g.

In aquaculture, most freshwater production of smolts is done in tanks or lakes (mostly Chile). Fish can either be maintained on ambient temperature and light regimes to produce 1½ year old smolts, or light and temperature regimes can be manipulated artificially to induce early smoltification at an age of 7-8 months. Production densities vary depending on the system; very intensive systems may maintain fish at densities as high as 50 kg/m³ or higher.

Smolted fish at 40-120 g are transferred to sea sites following once the smoltification process has taken place and the fish are adapted for seawater survival. Ongrowing at sea normally takes place in cages consisting of large nets suspended from various floating "walkway" systems anchored to the seabed, although some production has been carried out in pump-ashore seawater tank systems. Atlantic salmon grow best in sites where water temperature extremes are in the range 6-16 °C, and salinities are close to oceanic levels (33-34‰). Maximum stocking densities of up to 20 kg/m³ are usual. Atlantic salmon are ongrown in sea sites for 1-2 years with harvesting of fish from 2 kg upwards. Premature maturation may be a problem for both the size and quality of the product. Maturation results in loss of growth, browning of the external colour, decrease in fat, colour of the flesh and condition factor of the fish. Sea sites normally contain a single generation of fish. Good practice is to fallow sea sites for a period of 6 weeks or more prior to the introduction of a new generation of fish.

  • Generation interval for males and females

Atlantic salmon males and females can mature between 1 and 5 years, depending on strain and light and temperature regime. Most broodstock spawning is done at either 3, 4 or 5 years of age.


Broodstock are selected from sea site, and normally moved into freshwater tanks or cages in autumn prior to stripping. Salmon has a high fecundity with about 1 600 eggs per kg body weight. Eggs are stripped dry, fertilized with milt, water hardened and disinfected prior to transfer to trays or silo systems. Following eyeing, they are "shocked" to remove unfertilized eggs. Hatching takes place in hatchery trays or following transfer to tanks. Incubation of eggs and alevins normally takes place in water at <9 °C. First feeding, using inert feeds, is normally carried out following transfer of late alevins into tanks. "Feeding fry" can be grown on in tanks, either using flow-through or various recirculation systems, or subsequently in lake cage systems, through parr stages to smolt.
Main diseases

The rainbow trout is affected by several bacterial diseases: furonculosis (Aeromonas salmonicida), vibriosis (Vibrio spp.) and Bacterial kidney disease (Renibacterium salmoninarum). Effective vaccines exist for furonculosis and vibriosis.

Viral diseases are far more important economically in Atlantic salmon. The main viral diseases affecting Atlantic salmon are IPN (Infectious Pancreatic Necrosis), PD (pancreas disease) and ISA (Infectious salmon anaemia). The last one is a notifiable diseases in the EU Council Directive 91/67/EEC defining Approved Zones, which are free of such diseases.

Other major health problems affecting Atlantic salmon are sea lice (Caligidae) and fungal infections (Saproleignia spp.), especially during the freshwater rearing period and where juvenile salmon are reared in lakes.

1   2   3   4   5   6   7   8   9   ...   16

База данных защищена авторским правом © 2016
звярнуцца да адміністрацыі

    Галоўная старонка