Heritability of juvenile growth traits in red drum (Sciaenops ocellatus L.)

Heritability of juvenile growth rate was estimated for red drum (Sciaenops ocellatus), an economically important sciaenid fish in the southern USA. Thirty‐eight families were generated via ‘natural’ spawning of multiple sets of five breeders (three dams × two sires) in individual brood tanks. Offspring were individually tagged with Passive Integrated Transponder (PIT) tags and mixed for grow‐out in replicate ‘common‐garden’ tanks. Juvenile growth was followed from 166.4±18.6 to 254.0±27.0 mm (total length). Offspring were assigned a posteriori to individual brood fish (dam and sire) based on genotypes at nuclear‐encoded microsatellites. Heritability (h²) of a thermal growth coefficient was estimated using an animal‐additive model and a restricted maximum‐likelihood algorithm. Estimates of h² were 0.33±0.08 and 0.31±0.08 for thermal growth coefficient based on length and weight respectively. These results indicate a significant genetic component in juvenile growth rate in red drum. Estimates of h² for condition coefficient (K) at various measurement dates averaged 0.38, suggesting a genetic component to shape in juvenile red drum.     

Genetic parameters for resistance against Flavobacterium columnare in Nile tilapia Oreochromis niloticus (Linnaeus, 1758)

Columnaris disease is a major cause of mortality in tilapia hatcheries and commonly occurs during the summer season in Thailand. One way of reducing the problem is by selective breeding for increased disease resistance. The objective of this study was to estimate quantitative genetic parameters for resistance against columnaris in the Chitralada 4 strain of Nile tilapia. Data from 43 full‐sib families (2,580 records) of fry (age = 32 ± 4 days post‐hatch) were used in the analyses. Initially, fry were subjected to bath challenge with Flavobacterium columnare (LD₅₀ concentration = 1.2 × 10⁶ CFU/ml) for 14 days. Disease resistance was defined as the number of days from challenge until death (DD) or as a binary trait (dead/alive) on day 14. Linear animal and sire‐dam models were used for DD, while threshold animal, threshold sire‐dam, binary linear animal and binary linear sire‐dam models were used for binary outcomes. Covariate effect of age, fixed effect of challenge day and random effects of the individual animals or sires and dams were included in the models. Mean survival was 32.4 ± 11.6%, and survival rates of the best and poorest families were 70% and 8%, respectively. The highest estimate of heritability (0.30 ± 0.025) was obtained under the threshold sire‐dam model. Heritability estimates for DD (0.16 ± 0.034 and 0.17 ± 0.046) were comparable to those obtained from the threshold animal (0.15 ± 0.031) and the binary linear (0.14 ± 0.045 and 0.15 ± 0.044) models. The linear animal and sire‐dam models for DD and the threshold sire‐dam models performed equally with similar values of r_EBV (0.629, 0.628 and 0.627) and accuracy of selection (0.793, 0.793 and 0.791). This study reveals the potential of selective breeding to increase disease resistance to F. columnare in the studied population of Nile tilapia.     

Genetic effects on tolerance to acute cold stress in red drum,Sciaenops ocellatus L.

Genetic effects on cold‐stress tolerance were assessed for red drum (Sciaenops ocellatus), an economically important sciaenid fish in the southern USA. Thirty‐five families were generated via ‘natural’ spawning of multiple sets of five breeders (three dams × two sires) in individual brood tanks. Offspring from the 35 families were transferred abruptly from an acclimation temperature of ～24 °C to 5.7 °C and maintained subsequently at an average temperature of 6.5 °C. Offspring were assigned a posteriori to individual broodfish (dam and sire) based on genotypes at nuclear‐encoded microsatellites. Heritability of the survival–time probability function was estimated using a proportional hazard approach and an animal‐additive model. The estimated heritability was 0.20 (95% CI: 0.07–0.40), indicating a significant genetic component to acute cold‐stress tolerance in red drum.     

Genetic effects on carcass-quality traits in hybrid striped bass (Morone chrysops♀×Morone saxatilis♂)

A 10 × 10 factorial mating design was used to examine the genetic effects on various carcass‐quality traits in hybrid striped bass (Morone chrysops♀×Morone saxatilis♂). A total of 448 offspring were raised in a ‘common‐garden’ environment and carcass‐quality traits were assessed at 389 days post fertilization; parentage of each fish was inferred from genotypes at 5–10 nuclear‐encoded microsatellites. There was a significant effect of dam and sire on body weight and fillet weight and a significant effect of dam on viscera weight and condition factor. When carcass‐quality traits were standardized to body weight, there was a significant effect of dam, sire and dam × sire interaction on viscera percentage (proportion of viscera weight to body weight). Phenotypic correlations between body weight and non‐standardized carcass‐quality traits were high (r_p=0.80–0.92), whereas phenotypic correlations between body weight and standardized carcass‐quality traits were low (r_p=0.13–0.19). Genetic correlations between body weight and fillet weight (dams and sires), and body weight and viscera weight (dams only) were significant and positive, while the genetic correlations between body weight and viscera percentage (dams and sires) were significant and negative. Significant general combining ability values for favourable carcass‐quality traits in the same breeder (dam or sire) occurred only in one or two of 20 (dams and sires) used in this study, suggesting that simultaneous improvement in multiple traits in hybrid striped bass may require evaluation of large numbers of candidate breeders. The effect of sex on body weight and all non‐standardized carcass‐quality traits was significant and appeared to be due to faster growth in females. These results indicate that selection for faster growth of hybrid striped bass would lead to a correlated increase in fillet weight and viscera weight but not necessarily to an increase in the proportion of fillet and waste product generated. The significant dam, sire and dam × sire interaction effects on viscera percentage and the negative genetic correlations between body weight and viscera percentage indicate that selection for reduced proportion of viscera relative to body weight might be feasible.     

Estimates of Heritability of Larval and Early Juvenile Growth Traits in Red Drum (Sciaenops ocellatus)

ABSTRACT

Eighteen to 29 families of red drum were generated via spontaneous spawning of multiple sets of three dams?×?two sires. In 2002, offspring from spawning events were grown in separate larval ponds to a mean TL of 30.4 mm. In 2003, offspring from spawning events were individually passive integrated transponder (PIT)-tagged and grown in “common-garden” tanks from 121.9 to 166.6 mm. Offspring in both experiments were assigned to parents based on genotypes at four microsatellite loci. Heritability estimates were 0.24?±?0.06 (larval TL) and 0.48?±?0.16 (juvenile-specific growth rate in length) and indicate a significant genetic component for both traits.     

A genetic analysis of egg,alevin and fry mortality in salmon (Salmo salar), sea trout (Salmo trutta) and rainbow trout (Salmo gairdneri)

Heritability of mortality in eggs, alevins and fry were estimated using data from salmon, sea trout and rainbow trout. The highest value (average for all species and both estimation methods) based on the sire component was obtained for the eyed egg stage, h² = 0.08, followed by the alevin stage, h² = 0.05, while estimates for fry mortality did not differ significantly from zero.The maternal effect was rather large for uneyed eggs and eyed eggs, while it was low for alevin mortality. Maternal effects were also found to have a significant influence on the mortality of salmon fry. It was, however, not possible to separate maternal effects from tank effects.     

Phenotypic and genetic parameters of body size traits in Atlantic salmon Salmo Salar L.

Data on two year classes of Atlantic salmon, Salmo salar L., reared for 2 years in sea water, were analysed. Records were obtained on 1271 offspring of 58 sires and 171 dams mated hierarchically. Medium-high sire heritabilities (0.30-0.38) were found for ungutted and gutted body weights, body length, body circumference and condition factor (whole-body traits). Corresponding estimates for height and width of cutlet and average thickness of the abdominal wall ranged from 0.23 to 0.27. The latter measurements were made on a transversal section anterior to the dorsal fin. Heritability estimates derived from the dam component of variance corresponded in general to the sire estimates, indicating the absence of non-additive genetic and environmental effects common to full sibs. Estimated phenotypic and genetic correlations among the traits were all positive. With regard to the likely breeding goal of a commercial selection programme, the genetic correlations of ungutted and gutted body weights with condition factor (r_G=0.36-0.37) were judged unfavourable. The results revealed substantial impact of sex and sexual maturation status on growth. The need is stressed for proper adjustment for these effects when assessing growth or growth-related data.     

Phenotypic and genetic parameters of body composition traits and flesh colour in Atlantic salmon, Salmo salar L.

Data obtained on 3.5-year-old offspring of 58 sires and 171 dams (two year classes) of Atlantic salmon, Salmo salar L., were analysed. A substantial effect of sexual maturation status was demonstrated for several body composition traits, most pronounced for the content of fat and dry matter of muscle. Maturation effects generally exceeded estimated sex effects. Intermediate estimates of heritability (0.20-0.30) derived from the sire component of variance were obtained for intestinal fat weight, intestinal fat index, dressing percentage and fat and dry matter content of muscle. Corresponding parameters of the subjectively scored traits belly flap constitution, amount of intestinal fat and flesh colour were all low (= 0.10). Heritabilities derived from the sire and dam components of variance were generally similar in magnitude, indicating the absence of non-additive genetic and environmental effects common to full-sibs. Viewed in the context of likely selection goals for commercial breeding schemes, the results did not reveal major adverse genetic relationships among the investigated traits. However, the estimated genetic correlations of muscle fat with body weight (0.42 ± 0.17) and intestinal fat index (-0.64 ± 0.19) may be unfavourable.     

应用重复力模型估计虹鳟生长性状的遗传力和育种值

采用单性状重复观测值动物模型（重复力模型）对中国水产科学研究院黑龙江水产研究所渤海冷水性鱼类试验站5个品系后代的4026尾虹鳟 (Onchorynchus mykiss) 的生长性状（体重、体长和肥满度）进行了遗传参数和育种值的估计,并对各个性状的表型和遗传趋势进行预测。在此模型中考虑了年份-季节固定效应、父本固定效应和母本固定效应;随机效应有个体的加性效应和个体永久环境效应。结果表明：该5个品系虹鳟及后代体重的遗传力为0.35,体长为0.10、肥满度为0.34。个体间的育种值差异较大,仅排名前20位相差近10倍;从不同评定方法的秩相关来看,用综合育种值法与单个性状方法之间存在极显著的相关,其相关度分别为0.998, 0.877, 0.85. -0.071, -0.064和-0.13。用综合育种值对虹鳟个体的选择价值的评定和名次排列, 以及与单个性状的评定名次有着一定程度上的差别。体重、体长和肥满度性状,其表型趋势与遗传趋势基本一致的品系,均可通过表型值进行选择。     

Quantitative genetics and heritability of growth-related traits in hybrid striped bass (Morone chrysops ♀ × Morone saxatilis ♂)

Commercially farmed, hybrid striped bass - female white bass (Morone chrysops) crossed with male striped bass (Morone saxatilis) - represent a rapidly growing industry in the United States. Expanded production of hybrid striped bass, however, is limited because of uncontrolled variation in performance of fish derived from undomesticated broodstock. A 10 × 10 factorial mating design was employed to examine genetic effects and heritability of growth-related traits based on dam half-sib and sire half-sib families. A total of 881 offspring were raised in a common environment and body weight and length were recorded at three different times post-fertilization; parentage of each fish was inferred from genotypes at 10 nuclear-encoded microsatellites. Dam and sire effects on juvenile growth (weight and length) and growth rate were significant, whereas dam by sire interaction effect was not. The dam and sire components of variance for weight and length (at age) and growth rate were estimated using a Restricted Maximum Likelihood algorithm. Estimates of broad-sense heritability of weight, using a family-mean basis, ranged from 0.67 ± 0.17 to 0.85 ± 0.07 for dams; estimates for sires ranged from 0.43 ± 0.20 to 0.77 ± 0.10. Estimates of broad-sense heritability of growth rate (based on weight), using a family-mean basis, ranged from 0.69 ± 0.12 to 0.82 ± 0.09 for dams and from 0.69 ± 0.13 to 0.81 ± 0.08 for sires. Similar results were obtained with length data. Both genetic and phenotypic correlations between weight and length were close to unity. High genetic (0.98-0.99) and phenotypic (0.79) correlations between growth rates measured at two time intervals suggested that selection for growth rate at an early life stage could affect growth rate at a later life stage. Estimates of general combining ability (GCA) for growth rates differed significantly among dams and among sires, whereas estimates of specific combining ability (SCA) for each dam × sire combination did not differ significantly from zero. These results suggest that additive-effect genes contributed to the differences in juvenile growth.     

Genetic variations in survival of rohu carp (Labeo rohita, Hamilton) after Aeromonas hydrophila infection in challenge tests

A challenge test against Aeromonas hydrophila was conducted using 2062 rohu carp (Labeo rohita) fingerlings obtained from 52 dams and 87 sires (87 full-sib families) of two year-classes (2003 and 2004). Attempts to establish a cohabitant challenge model were not successful. Therefore, fish were challenged by intraperitoneal injection with A. hydrophila in two replicate tanks per year-class, and dead fish were collected hourly. The mortality reached its peak at 16–22 h after challenge and had almost completely stopped after 58 h. The test was terminated after 382 h at which the average survival was 43.9 and 48.9% in the two 2003 year-class tanks, and 77.1 and 35.8% in the two 2004 year-class tanks. Heritability estimates for survival were obtained from sire and dam threshold models (THR) and sire and dam linear repeatability models (LINR). For both types of models the heritability estimates based on data from single tanks were not consistent. For both year-classes, data from one of the two challenged tanks demonstrated significant additive genetic variation in survival during the A. hydrophila infection, whereas the heritability estimates were not significantly different from zero for the other tank. Further, genetic correlation between survivals in the two replicate tanks in each year-class was not significantly different from zero. The differential results from the replicate tanks demonstrate that additional challenge test experiments are needed before firm conclusions can be drawn about the magnitude of additive genetic variation for survival to aeromonasis in rohu carp. A cohabitant challenge model that allows the testing of important defence mechanisms in the skin and mucous membranes of the fish might have been more appropriate. To establish a valid cohabitant challenge model for rohu carp should be given high priority.     

Covariation between diploid and triploid progenies from common breeders in rainbow trout, Oncorhynchus mykiss (Walbaum)

The covariation between diploid and triploid progenies from common breeders was investigated in various progeny-testing experiments where either dams or sires were sampled from rainbow trout, Oncorhynchus mykiss (Walbaum), stocks. Triploidiza tion was found to frequently reduce the performance in the traits studied: body length and weight, growth, coefficient of condition and pyloric caeca number. Triploidization also generated significant interactions with the parental breeding value. These interactions were caused in part by the familial variance not being the same in triploids as in diploids, but also by actual ranking differences between diploid and triploid familial performances. However, the effect of these interactions was minor as compared with the amount of variation common to both ploidy levels (genetic correlations averaged 0.7–0.9). Therefore, selection of diploid breeders appeared efficient enough for improving triploid progeny, unless family selection methods including triploid progeny testing were preferred for other reasons. Lastly, it was observed that variances from maternal origin tended to be larger in triploids, whereas variances from paternal origin tended to be smaller, as compared with diploids. This point was discussed referring to the genetic make-up of triploids and in the scope of dams and sires for selective breeding.     

Effects of feed processing method (extrusion and expansion-compression pelleting) on water quality and growth of rainbow trout in a commercial setting

To determine the effects of feed processing method (extrusion or expansion followed by compression pelleting) on feed physicochemical characteristics, fecal stability, water quality, and growth performance in rainbow trout, three types of feed pellets (expanded sinking [EpS], extruded sinking [ExS], extruded floating [ExF]) were prepared, analyzed, and fed to juvenile rainbow trout (initial weight = 285.8 ± 15.2 g) at three feeding rates based on feeding rate curves developed at Snake River Farm (Buhl, Idaho, USA) (median rates [%BW] for feeding rate 1: 0.85; feeding rate 2: 0.95; feeding rate 3: 1.05) for 124 days. Results showed that all feeds had similar chemical composition, but the degree of starch gelatinization (as % of starch) was significantly higher in ExS (90.01 ± 0.26) and ExF (84.82 ± 0.63) than in the EpS (9.09 ± 0.41) feed, which led to extruded feeds having much higher water stability, fecal durability, and lower phosphorus discharge. The ExF feed produced better growth (% increase; 154.4 ± 18.5) and feed conversion (1.06 ± 0.06) compared to EpS (126.3 ± 12.4 and 1.26 ± 0.14) and ExS (121.7 ± 7.8 and 1.22 ± 0.04) in rainbow trout. Growth (% increase) improved as feeding rate increased (feeding rate 1: 125.6 ± 10.3; feeding rate 2: 140.8 ± 16.0), but further improvements were not observed at the highest feeding rate (feeding rate 3: 135.9 ± 22.8). The present study is the first to show that the extrusion process not only produces pellets having better quality than the expansion-compressed pelleting method but also improves fecal size and durability in water, and therefore, use of extruded feeds has the potential to improve waste collection and removal and reduce contribution to pollution in effluent.     

Effects of oral passive immunization against somatostatin‐14 on growth performance,body composition and IgY delivery in rainbow trout (Oncorhynchus mykiss) and common carp (Cyprinus carpio)

Egg‐yolk antibodies (IgY) against somatostatin‐14 (anti‐SST‐14) were evaluated as orally administered, growth promotants in gastric rainbow trout (Oncorhynchus mykiss) and agastric common carp (Cyprinus carpio) in an 8‐week feeding trial. Feeding groups were compared with fish which did not receive anti‐SST‐14 IgY. Growth responses and IgY plasma contents of the blood were assessed. In contrast to rainbow trout, oral anti‐SST‐14 addition to carp significantly improved protein efficiency ratio (PER, 1.02 ± 0.04) and protein productive value (PPV, 26.7 ± 0.91) after 56 days (P < 0.05) compared to the negative control (PER = 0.91 ± 0.02; PPV = 22.9 ± 0.66). IgY was undetectable in plasma of rainbow trout after oral administration, indicating that gastric degradation of the fed IgY makes this application route for growth promotion in gastric trout challenging.     

Effect of triploidy on paternal and maternal variance components in brown trout, Salmo trutta L.

The covariation between diploid and triploid progenies from common breeders and the effect of triploidy on the parental variances were investigated in brown trout (Salmo trutta L.) using two progeny testing experiments, sampling, sires and dams respectively, from the same population. The traits studied were body weight, growth, condition factor and red spotting of the skin. Triploidization generated some interactions with the parental breeding value, but their effect was minor (less than 20% of the genetic variance, in most cases) as compared with the amount of variation common to both ploidy levels. These interactions were mainly caused by a scale effect, triploidy reducing the variance attributable to sires and increasing the variance attributable to dams. Actual lack of correlation (genetic correlation significantly less than 1) between diploid and triploid familial performances was observed in a single instance out of 18. The modification of respective parental variance components by triploidy, already observed in the rainbow trout Oncorhynchus mykiss (Walbaum), appears as a logical consequence of the genetic make‐up of triploids, and should be taken into account in selective breeding.     

A riverscape perspective of Pacific salmonids and aquatic habitats prior to large‐scale dam removal in the Elwha River,Washington, USA

Abstract Dam removal has been increasingly proposed as a river restoration technique. In 2011, two large hydroelectric dams will be removed from Washington State’s Elwha River. Ten anadromous fish populations are expected to recolonise historical habitats after dam removal. A key to understanding watershed recolonisation is the collection of spatially continuous information on fish and aquatic habitats. A riverscape approach with an emphasis on biological data has rarely been applied in mid‐sized, wilderness rivers, particularly in consecutive years prior to dam removal. Concurrent snorkel and habitat surveys were conducted from the headwaters to the mouth (rkm 65–0) of the Elwha River in 2007 and 2008. This riverscape approach characterised the spatial extent, assemblage structure and patterns of relative density of Pacific salmonids. The presence of dams influenced the longitudinal patterns of fish assemblages, and species richness was the highest downstream of the dams, where anadromous salmonids still have access. The percent composition of salmonids was similar in both years for rainbow trout, Oncorhynchus mykiss (Walbaum), coastal cutthroat trout, Oncorhynchus clarkii clarkii (Richardson) (89%; 88%), Chinook salmon, Oncorhynchus tshawytscha (Walbaum) (8%; 9%), and bull trout, Salvelinus confluentus (Suckley) (3% in both years). Spatial patterns of abundance for rainbow and cutthroat trout (r = 0.76) and bull trout (r = 0.70) were also consistent between years. Multivariate and univariate methods detected differences in habitat structure along the river profile caused by natural and anthropogenic factors. The riverscape view highlighted species‐specific biological hotspots and revealed that 60–69% of federally threatened bull trout occurred near or below the dams. Spatially continuous surveys will be vital in evaluating the effectiveness of upcoming dam removal projects at restoring anadromous salmonids.     

Effect of temperature and feed ration on energy utilization in large rainbow trout, Oncorhynchus mykiss (Walbaum)

Abstract In one experiment, temperature and feed ration were varied for rainbow trout, Oncorhynchus mykiss (Walbaum), with an initial size of about 1200g. In two experiments the feed ration was varied between parallel tanks, with naturally fluctuating ambient temperatures, and containing rainbow trout with an initial size of about 1000g and 400–550g, respectively. Growth and feed efficiency were measured in all experiments. An ad libitum feed ration without wastage produced the most effective energy utilization in most experimental groups. At best an average of 50% of the gross energy was utilized for growth. The relation between growth rate and feed ration was linear, and there was no optimum point for feeding below this level. The observed intake of energy is compared with growth, measurements of digestibility, feeding and starving metabolism. It is also linked to earlier observations of intake by rainbow trout fingerlings to give equations expressing the energy needs of rainbow trout of different sizes (30–1500g). These are intended to be used for efficient rearing at different temperatures (10–22°C).     

Growth performance,nutrient digestibility and intestinal morphology of rainbow trout (Oncorhynchus mykiss) fed graded levels of the yeasts Saccharomyces cerevisiae and Wickerhamomyces anomalus

In a 10‐week study, we evaluated the effects of replacing 20%, 40% or 60% of fishmeal (present in control diet at 300 g/kg) on a digestible protein basis with yeast Saccharomyces cerevisiae or a yeast mixture of Wickerhamomyces anomalus and S. cerevisiae on growth performance, nutrient digestibility, nutrient retention and intestinal health of rainbow trout (Oncorhynchus mykiss). Triplicate tanks with 35 rainbow trout (144.7 ± 25.1 g mean ± SEM) were fed rations of 1.5% of total biomass per tank. Replacement of 60% of fishmeal with yeast mixture resulted in lower specific growth rate of 1.0% versus 1.2%/day for other diets. Apparent digestibility coefficients for crude protein and most amino acids were highest in fish fed fishmeal‐based diet, with similar values for fish fed the diet with 20% replacement with yeast mixture. Diet with 20% replacement with yeast mixture resulted in highest phosphorus digestibility. Replacement of 60% of fishmeal with S. cerevisiae resulted in oedematous mucosal fold tips in the proximal intestine. The results of this study suggest that these yeasts can replace up to 40% of fishmeal under current inclusion levels in diets for rainbow trout without compromising growth performance, nutrient digestibility or intestinal health.     

Genetic improvement of Israeli blue (Jordan) tilapia,Oreochromis aureus (Steindachner), through selective breeding for harvest weight

Genetic parameters and selection responses were obtained for harvest body weight of blue tilapia (Oreochromis aureus) from data collected over three generations in a selected population. A total of 18 194 records representing 186 sires and 201 dams were used in the analysis. Within generation heritability estimates for harvest body weight ranged from 0.18 to 0.58. When data from more than one generation were included in the analysis, heritability estimates became more stable (0.33–0.40) and it was 0.33 when all data were included in the analysis. The common full‐sib effect accounted for 10% of the phenotypic variance in the full data set. Heritability for survival from stocking to harvest was estimated at 0.01 and 0.09 in actual units (fitting an animal model) and in the logit (sire model) scale respectively. The genetic correlation between harvest body weight and survival was 0.22 and not significantly different from zero. The total selection response for harvest body weight over the three generations of selection measured as the difference between least‐squares means of selected and control lines was 17.7%. The corresponding figure when response was measured as the difference between mean breeding values of selected and control lines was 19.6%. The average inbreeding coefficient was 0.003 after three generations of selection. These results indicate that there are good prospects for the genetic improvement of harvest body weight in blue tilapia.     

Paternal variation in juvenile survival and growth of the triploid hybrid between female rainbow trout (Oncorhynchus mykiss Walbaum) and male brown trout (Salmo trutta L.)

Abstract Related sib‐groups of rainbow trout × brown trout triploid hybrid and monospecific brown trout controls were obtained from a common set of brown trout sires. On the basis of hybrid juvenile survival and growth, 10 sib‐groups were selected and the corresponding brown trouts were raised up to adult stage. Males from each group of brown trout were used to produce a second generation of hybrid progeny, the performances of which were analysed for grandpaternal variation and relation with first generation relatives. Results showed that hybrid traits were influenced by paternal and grandpaternal origin, but that there was little relation between hybrids and brown trouts, or between successive generations of hybrids.