Salmon muscle growththe myosepta was slightly more advanced in triploid than in diploid ﬁsh, although the differences were smaller than among individual families. The c-met receptor

Polyploidy has been a significant factor in the evolution ofsalmonids (Allendorf and Thorgaard, 1984; Johnson et al.,1987). Low frequencies of triploid salmonids occur in naturalpopulations (Thorgaard and Gall, 1979), and triploidy can bereadily induced experimentally by either heat or pressuretreatment (Benfey, 1991). There are fewer, but larger, cells inmost organs and tissues of triploid fish, and it has long beensuspected that this may have far-reaching consequences fortheir development and biology (Small and Benfey, 1987;

Benfey, 1991). For example, the lower numbers of brain andsensory cells in triploids is correlated with marked changes inbehaviour, including a decreased sensitivity to light and sound(Aliah et al., 1990) and reduced aggressiveness relative todiploids (Benfey, 1991). Although mitotic cell divisionsproceed normally in triploids, the presence of the thirdchromosome means that balanced chromosome constitutionscannot be established at first meiosis. Triploids are thereforefunctionally sterile. In fish farming, maturation essentially

1991The Journal of Experimental Biology 202, 1991–2016 (1999)Printed in Great Britain © The Company of Biologists Limited 1999JEB1947

Muscle development and growth were investigated indiploid populations of normal-sex-ratio and all-femaleAtlantic salmon (Salmo salar L.) and their triploidcounterparts produced by high-pressure treatment.Somites were formed at the rate of 6 h−1 in both diploidsand triploids at 6 °C. The rostral-to-caudal development ofmyotubes, myofibrils and acetylcholinesterase staining atthe myosepta was slightly more advanced in triploid thanin diploid fish, although the differences were smaller thanamong individual families. The c-met receptor tyrosinekinase was used as a molecular marker for the satellite cellsinvolved in postembryonic muscle growth. Satellite cellnuclei comprised 17.5 % of total myonuclei in smolts andthey were 24 % more abundant in diploid than in triploidfish. Cells expressing the myogenic regulatory factor myf-6, a marker of satellite cells committed to differentiation,represented 14.8 % of total myonuclei in diploids and12.5 % in triploids. At ambient temperatures, the numberof white muscle fibres in normal-sex-ratio fish increasedmore than 30-fold between the alevin and smolt stages, andapproximately 3.5-fold further during the first year ofseawater growth. The rate of muscle fibre recruitment inseawater stages was significantly greater in diploidthan in triploid fish, reaching 1162 fibres day−1 and608 fibres day−1, respectively, in all-female groups 800 dayspost-hatching. For 42 cm fork-length fish, there wereapproximately one-third more muscle fibres per myotomein diploid than in triploid groups, 649 878 and 413 619,

respectively, for all-female fish. The probability densityfunction of muscle fibre diameters in each fish wasestimated using non-parametric smoothing techniques, andthe mean densities for diploids (fD) and triploids (fT) werecalculated. The peak fibre diameter was approximately20 µm in all age classes, irrespective of ploidy. Distinctbimodal distributions of muscle fibre diameter wereevident in all groups 775 days and 839 days post-hatching,reflecting seasonal cycles of fibre recruitment. fD and fT

were compared using a non-parametric bootstraptechnique and the reference band representing the null-hypothesis indicated that there was no difference withploidy. Reference bands for normal-sex-ratio fish at 315days and 470 days indicated that diploids had a higherpercentage of smaller-diameter fibres and that triploiddistributions had a thicker right-hand tail. Similardifferences in fD and fT of muscle fibre diameters werefound for all-female fish, although the statistical evidencewas less strong. Reference bands indicated differences inthe middle range of the distributions of muscle fibrediameter in fish 620–775 days post-hatch, with triploidshaving a thicker right-hand tail. Thus, a lower density ofsatellite cells was associated with reduced rates of fibrerecruitment but a compensatory increase in muscle fibrehypertrophy in triploid compared with diploid fish.

Key words: Atlantic salmon, Salmo salar, muscle, growth,development, ploidy, sex-reversed fish.

Summary

Introduction

MUSCLE GROWTH AND DEVELOPMENT IN NORMAL-SEX-RATIO AND ALL-FEMALE DIPLOID AND TRIPLOID ATLANTIC SALMON

IAN A. JOHNSTON1,*, GILLIAN STRUGNELL1, MARTI L. MCCRACKEN2 AND RAY JOHNSTONE3

1Gatty Marine Laboratory, School of Environmental and Evolutionary Biology, 2School of Mathematics andComputational Sciences, University of St Andrews, St Andrews, Fife KY16 8LB, Scotland and 3Fisheries Research

Service, Marine Laboratory, SOAEFD, Aberdeen, Scotland*e-mail: [email protected]; http://www.st-and.ac.uk/~fmrg/

Accepted 24 April; published on WWW 7 July 1999

1992

marks the end of the useful period of rearing and, if economicyields are to be maximised, fish must be sold before thedeteriorative changes associated with maturation lower theirmarket value. Salmon farmers became interested in the use oftriploid fish for sterility reasons (for a review, see Ihssen et al.,1990). Because of the potential genetic introgression threatposed by the escape of diploids from fish farms, salmonid stockmanagers are also interested in the sterility of triploids (for arecent review, see Youngson et al., 1997). If triploid salmoncould be reared as economically as diploids, then theintrogression threat to wild populations would be minimised atlittle cost to farmers.

Although male triploids are functionally sterile, theyproduce near-normal levels of circulating androgens atspawning time and therefore suffer the normal deteriorativechanges associated with maturation (Lincoln and Scott, 1984).Female triploids, because of differences in the architecture ofthe germinal and endocrine tissues, are both functionally andhormonally sterile (Sumpter et al., 1984). Only triploid femaleswould therefore be of use in aquaculture. Sex in fish, althoughdetermined by the presence of a Y male-inducing chromosome,is extremely susceptible to manipulation. The induction offunctional testes in genetic females (XX) is thereforestraightforward, involving the addition of small amounts ofhormones to the feed or rearing water (Johnstone andYoungson, 1984). These ‘inverted females’ (sex-reversedmales) can be used to fertilise normal eggs (XX) to produceall-female offspring from which the Y chromosome has beeneliminated.

The effects of triploidy on growth are variable and probablydepend on the induction method, husbandry practices and thestage of the life cycle being compared (Thorgaard, 1986).Several studies have reported that triploid fish grow moreslowly than diploids when reared under communal conditions,which may be related to decreased aggressiveness and/or anincreased susceptibility to stress. However, triploid Atlanticsalmon often show growth performance as good as or betterthan that of diploids when reared separately (Carter et al.,1994; Johnstone et al., 1991).

Somatic growth is closely related to that of the muscle tissue,which comprises approximately 65 % of the body mass(Weatherley et al., 1979). In Atlantic salmon, distinct germinalzones of myoblasts are present at the dorsal and ventral apicesof the myotome during the yolk-sac stages (Higgins andThorpe, 1990; Johnston and McLay, 1997). By first feeding,there are approximately 10 000 white muscle fibres permyotome, and at this stage small-diameter muscle fibres beginforming throughout the myotome on the surface of existingfibres (Higgins and Thorpe, 1990; Johnston and McLay, 1997).Surprisingly, given the importance of Salmo salar as a foodfish, there have been no systematic studies of musclerecruitment during the seawater stages of the life cycle,although it is known that there are in excess of 1 million musclefibres per myotome after two sea-winters of growth (Johnston,1999).

Muscle is a post-mitotic tissue, and post-embryonic growth

involves the satellite cell population, which proliferates toprovide a source of nuclei for fibre recruitment andhypertrophy (Veggetti et al., 1990; Koumans et al., 1991;Johnston et al., 1995, 1998). It has been reported that the yieldof mononuclear cells per gram of muscle tissue is significantlyhigher in primary cell cultures from diploid than it is fromtriploid rainbow trout (Oncorhynchus mykiss) (Greenlee et al.,1995), suggesting that there may be more satellite cells indiploids which, in turn, would be predicted to affect musclegrowth characteristics. However, in rainbow trout greater than3 cm in fork length, Suresh and Sheehan (1998) found that thefrequency histograms of muscle fibre diameter were similar indiploid and triploid fish. These authors indirectly estimated thatthere were 10 % fewer muscle fibres per unit cross-sectionalarea in triploid than in diploid individuals.

The objective of the present study was to provide acomprehensive description of muscle growth in diploid andtriploid Atlantic salmon (Salmo salar L.) using both normal-sex-ratio and all-female populations. Specifically, we wishedto test the hypotheses that ploidy influences the relative timingof embryonic myogenesis, the number of satellite cells permyotome and the rate of muscle fibre recruitment. Non-parametric smoothing techniques were applied in a novelapproach to investigating the density distribution of musclefibre diameter, and hence hypertrophic growth, at differentstages of the life cycle.

Recently, the c-met tyrosine kinase receptor has beenidentified as a molecular marker of satellite cells in the mouse(Cornelison and Wold, 1997). The MyoD family of myogenicregulatory factors (MyoD, myogenin, myf-5 and myf-6) playsan important role in muscle differentiation and the factors areexpressed in proliferating satellite cells (Grounds et al., 1992;Yablonka-Reuveni and Rivera, 1994; Cornelison and Wold,1997). In the present study, the use of antibodies to c-met andmyf-6 proteins to estimate the number of satellite cells insalmon smolts is also reported.

Materials and methodsExperimental stocks

Three experimental series were conducted. In the first series,beginning in 1994, the eggs from two females taken from a fishfarm were pooled and fertilised with either the milt from anormal farmed male, thus producing a population with anormal sex ratio (NSR), or with the milt from a sex-invertedfemale, thus producing an all-female population (AF; forexperimental details, see Johnstone and MacLachlan, 1994).Shortly after fertilization, a proportion of both egg batches wasmade triploid by high-pressure treatment (5 min at6.9×104 kPa) (Johnstone and Stet, 1995), thereby generatingfour comparative groups: diploid NSR, diploid AF, triploidNSR and triploid AF. This series of fish was reared for aprolonged period at the Fisheries Research Service salmon-rearing facility at Aultbea, Wester Ross, Scotland, at ambient,i.e. fluctuating, temperature. Fish were weaned onto dryproprietary feed when the yolk was approximately 80 %

I. A. JOHNSTON AND OTHERS

1993Salmon muscle growth

exhausted (first feeding). S1 parr destined to become smoltsafter 1 year in fresh water were selected in September 1995.Fish were transferred to tanks at a uniform stocking densityand fed with automatic feeders. All fish feeds were suppliedby Trouw Aquaculture Ltd. Fry were initially fed unpigmenteddry pellets 0.3–0.8 mm diameter at the rate of3 % kg−1 biomass day−1. Pellet size was increased with growth,and feeding rates varied from 2 to 2.55 % kg−1 biomass day−1

during the summer months to 1 % kg−1 biomass day−1 in thewinter. Fish were transferred to sea water in April 1996 andreared in replicated tanks at comparable stocking densities untilMay 1997. The temperature was recorded daily throughout theexperiment. The minimum and maximum temperaturesexperienced by freshwater stages were 1.7 °C and 21 °C,respectively, compared with 8 °C and 14 °C for seawater fish(Fig. 1).

In the second series of experiments, started in 1995, eggsfrom two wild salmon were fertilized with the milt from eitheran NSR farmed male or a sex-inverted female to produce twofamilies of NSR and AF offspring. Approximately half theeggs were made triploid as before, and both groups wereincubated at a constant temperature of 6±0.2 °C until theyhatched. In the third experimental series (in 1996), the eggsfrom six female wild salmon were fertilized with the milt fromone of two farmed males to produce six unique families of NSRfish. Diploid and triploid (prepared by pressure treatment asdescribed above) groups were reared until first feeding at aconstant temperature of 6±0.2 °C at the SOAEFD hatcheryfacility at Almondbank, Perthshire, Scotland.

Using the protocol of Johnstone and Stet (1995),triploidisation success was 100 % in all three experimental

series as monitored by light microscopic analysis of musclenucleoli (diploids had one or two and triploids one, two or threeper nucleus) or, in larger fish, by flow-cytometric analysis oferythrocytes.

Embryonic development

Embryos from each family were sampled daily until the endof somitogenesis and every 2 or 3 days thereafter. Eggs weredechorionated using fine forceps, and the embryos wereremoved without damage. From each sample, at least six livetriploid and six live diploid embryos were examined from eachfamily using a stereo microscope under both bright- and dark-field illumination. Once movements had started, the embryoswere anaesthetized in a 1:5000 (m/v) solution of bicarbonate-buffered MS222 (ethyl m-aminobenzoate). The somite stageand the appearance of particular organs and tissue types werenoted. Approximately 10 embryos from each group were fixedin each of the following fixatives (A) Bouin’s fluid and (B) 4 %(m/v) paraformaldehyde in 0.12 mmol l−1 phosphate-bufferedsaline (PBS) for subsequent histological and histochemicalanalysis, respectively. Embryos fixed in Bouin’s fluid wereembedded in wax, and 7 µm serial sagittal sections were cutand stained with haematoxylin–eosin. The most posteriorsomite containing myotubes and myofibrils was scored foreach embryo along with the somite stage. The appearance offunctional endplates at the myosepta was investigated bystaining embryos for acetylcholinesterase activity. Embryoswere incubated in the dark for 3–5 h at 4 °C in a solutioncontaining (in mmol l−1): copper sulphate, 3; potassiumferricyanide, 0.5; maleate buffer, 100; acetylthiocholine, 1.7.The staining reaction was stopped by rinsing several times inPBS, and embryos were mounted in glycerol under glasscoverslips supported by silicone grease at each corner andexamined using Nomarski differential interference (DIC)optics with a Leitz DRM Systems microscope.

Studies of muscle growth

Fish were sampled 46 and 57 days after fertilization (embryostages), at hatching and at first feeding, and fixed in Bouin’sfluid prior to processing for wax histology. Following firstfeeding, muscle growth was assessed using frozen sections toavoid problems associated with shrinkage. A shrinkagecorrection factor was applied to fibre size measurements of theearlier stages on samples fixed in Bouin’s fluid. To quantifymuscle cellularity, a cross section 3–5 mm thick was cut at thelevel of the pelvic fin insertions and photographed againstgraph paper. The cross section from one half of the body wasdivided into 3–10 labelled blocks depending on the size of thefish. Blocks were mounted on cork strips and frozen in 2-methyl butane cooled to near its freezing point (−159 °C) inliquid nitrogen. Samples were wrapped in tin-foil to avoiddesiccation and stored in a liquid nitrogen refrigerator untilthey could be processed. Frozen sections, 10 µm thick, werecut, air-dried and stained with the nuclear stain Scarba Red.The outlines of muscle fibres and the total cross-sectional areaof muscle tissue were digitized using an image-analysis

Age post-hatch (days)

Tem

pera

ture

(°C

)

0

5

10

15

20

25

FF

SWT

S1

0 100 200 300 400 500 600 700 800

Fig. 1. Water temperatures recorded during 1994–1997 at the Altbeafish cultivation unit, West Rosshire, Scotland. The filled symbolsrepresent the temperatures for the freshwater stages (triangles,embryonic; circles, post-hatch) and the open symbols for thesaltwater stages of the Atlantic salmon (Salmo salar L.) familiesstudied. Although the temperature was recorded daily for clarity,only the data for approximately every tenth day are shown. FFcorresponds to first feeding, S1 to the selection of fish in the uppergrowth mode during summer 1995 and SWT to the transfer of smoltsto saltwater tanks.

1994

system, and the equivalent muscle fibre diameters werecalculated (Kontron Electronics, Basel, and ScanBeam,Denmark). The estimated value of the total number of musclefibres per myotome was plotted against the cumulative numberof fibres sampled until a stable estimate was obtained. Between800 and 1200 muscle fibres were measured per fish,representing fields from all areas of the myotomal crosssection.

Immunohistochemistry

Transversely cut 7 µm thick frozen sections of muscle weremounted on poly-L-lysine-coated glass slides and air-dried forapproximately 1 h. Sections were fixed for 10 min in 4 % (m/v)paraformaldehyde in PBS, washed three times for 3 min inPBS, blotted and placed in acetone for 10 min and finally air-dried for 10 min. Prior to immunohistochemistry, sections wererehydrated in 1 % (v/v) Triton X-100 and 1 % (m/v) bovineserum albumin (BSA) (Sigma Chemicals, Poole, UK) in PBS.Background peroxidase activity was reduced by incubating thesections in 0.5 % (v/v) hydrogen peroxide for 10 min, and non-specific binding sites were blocked for 15 min with a solutioncontaining 4 % (m/v) normal goat’s serum, 1 % (v/v) Triton X-100 and 1 % (m/v) BSA in PBS. Sections were incubatedovernight in the primary antibody at 4 °C. Primary antibodiesand dilutions used were as follows: rabbit anti-m-met (SantaCruz) at 1:100, which stains the c-met tyrosine kinase receptor(Cornelison and Wold, 1997); and rabbit anti-myf 6 at 1:100(Santa Cruz). After three washes in PBS, sections wereincubated for 1 h in the secondary antibody biotinylated goatanti-rabbit IgG (Sigma Chemicals, Poole, UK) at a dilution of1:20 in 1 % (v/v) Triton X-100, 1 % (m/v) BSA in PBS. Twocontrol incubations were carried out omitting either theprimary or the secondary antibody. Sections were washed threetimes in PBS and incubated for 1 h in a 1:20 dilution ofextraAvidin peroxidase (Sigma Chemicals, Poole, UK) in 1 %(v/v) Triton X-100, 1 % (m/v) BSA in PBS. Peroxidase activitywas developed using 3-amino-9-ethylcarbazole, which gives ared insoluble end-product. Duplicate sections werecounterstained in Mayer’s haematoxylin to visualise totalmyonuclei. Slides were mounted under coverslips usinggelatine and stored in the dark. Counts of the number of c-met-/myf-6-positive cells per nucleus were made from at least sixfields of approximately 50 muscle fibres per field in each fishusing an image-analysis system (Scanbeam Ltd, Denmark) andrelated to the total cross-sectional area of muscle.

Electron microscopy

Small bundles of white muscle fibres were isolated from thedorsal epaxial myotomes of NSR salmon at first feeding andfixed overnight in 2.5 % (v/v) glutaraldehyde, 2.5 % (m/v)paraformaldehyde in 100 mmol l−1 sodium cacodylate buffer,pH 7.4 at 4 °C. Samples were processed for electronmicroscopy as described previously (Johnston et al., 1995).Sagittal ultrathin sections of muscle fibres were cut, stainedwith lead citrate and uranyl acetate, and viewed with a Philips301 transmission electron microscope. The dimensions of

satellite cells were measured from photomicrographs at amagnification of 5000 times.

Statistics

The relationship between developmental characters and ageor somite interval was fitted by linear least-squares regression.A two-way general linear model (GLM) analysis of covariance(ANCOVA) was used to test for differences in development,with ploidy and family as fixed effects and age or somiteinterval as a covariate (SPSS Statistical Software, SPSS Inc.,USA). The somite interval (the mean time to make a somite)was calculated by least-squares regression and is equivalent tothe somite stage of the embryo up until the end of somiteformation. Growth variables such as body mass, the total cross-sectional area of muscle and the number of fibres per myotomewere plotted against either age post-hatch or fork length. Datawere fitted by least-squares regression using a second-orderpolynomial and, in cases where a rate variable was required,the resulting curve was differentiated (Mathematica software,Wolfram Research Inc., USA). The hypothesis that there wasa single population underlying the fitted curves was testedusing multiple regression analyses (Zar, 1984). The effects ofploidy on muscle cellularity variables was tested usingANCOVA with fork length as a covariate (SPSS statisticalsoftware, SPSS Inc., USA). Satellite cell densities werecompared using a one-way analysis of variance (ANOVA).The data for NSR and AF fish were analysed in separateANCOVAs since these groups were half-siblings and thereforediffered in genetic background.

To evaluate and compare distributions of muscle fibrediameter, smooth non-parametric estimates of the probabilitydensity functions (PDFs) were constructed using the kernelapproach (Silverman, 1986) within the S-Plus computingenvironment utilising the sm library (Bowman and Azzalini,1997). The kernel method uses a smooth kernel function thatis itself a PDF, such as the normal curve, as the basic buildingblock. These smooth functions are centred directly over eachobservation, resulting in a smooth estimate of density whilepreserving the value of the realisation. The variance of thekernel function is controlled by the smoothing parameter h.The kernel estimator used was of the form:

where f̂ is the estimated probability density function, yi is theith observation from the list of n and w is the kernel function.

For simplicity when comparing groups, we sampled an equalnumber n of fish in each group, and we sampled an equalnumber m of fibres within each fish. We then estimated thePDF for each fish using the normal optimal smoothingparameter (Bowman and Azzalini, 1997). Sample sizes were800–1000 per fish, and smoothing parameters were within therange 0.13–0.19. The mean PDFs for diploid (fD) and triploid(fT) groups were then estimated using the diameters pooled

(1)f̂ (y) = w(y − yi;h) ,^n

i=1



over group. Since it is advantageous to use a commonsmoothing parameter when comparing densities (Bowman andAzzalina, 1997), fD and fT were estimated with h equal to themean of the normal optimal smoothing parameters over fishwithin an age class. To restrict diameters to positive values, weestimated density functions of the natural logarithm ofdiameter and then transformed back to the original scale.Furthermore, we fixed the maximal diameter within an ageclass at 110 % of the maximum diameter in the age class. Sincethe fit of the right-hand tail is dependent on the right endpointof possible diameters, fixing the maximal diameter wasrequired for consistency when comparing tail percentages. Fig.2 shows a histogram and mean density estimate of fibrediameters for an NSR fish at 775 days post-hatch as anexample.

Approximating confidence bands for fD and fT would be thenext logical step; however, the kernel estimator is biased, andthe form of this bias as well as the form of the variance of f̂makes approximating confidence bands computationally verycomplicated (Bowman and Azzalina, 1997). An alternative toa confidence band is a variability band. A variability bandrepresents only the variability of an estimate, but it is stillvery helpful in strengthening the evidence of a structurecharacteristic, such as a right-hand tail or a bimodaldistribution. The variability band of f̂ was approximated usingbootstrap techniques.

The hierarchical structure of the data (the observational unitwas fish, with fibres being subunits) results in two sources ofvariation: variation among fish and variation within fish. It isimportant to take into account these two sources of variationwhen drawing bootstrap samples. One strategy is to samplerandomly n fish with replacement and then to sample randomly,with or without replacement, m fibres within a fish. For amoderate sample size of, for example, 10 fish, sampling fibreswithout replacement is preferred (Davison and Hinkley, 1997).Since we had a small sample size of fish, we selected fibreswithin a fish using a smooth bootstrap sample as described bySilverman (1986) and Davison and Hinkley (1997).

For each bootstrap sample, the mean density was estimated

and the area between this estimate and f̂ D or f̂ T was shaded.The final shaded area represents the maximal area created bythe 100 bootstrap estimates of density and is referred to as thevariability band. If the structural characteristic of interest isalso common in the area, there is evidence that thecharacteristic is genuine. For example, if after 100 bootstrapreplications the area did not have the characteristics of abimodal distribution apparent in the estimate of mean density,we concluded that there was insufficient evidence to support abimodal distribution.

The next step of the analysis was to test the null hypothesisthat fD=fT over all diameters. To test this hypothesis, non-parametric bootstrap test procedures were used. First, theKolmogorov–Smirnov two-sample test statistic Dmn wascalculated:

where Smn(y) and Tmn(y) are the empirical distributionfunctions of the m↔n fibre diameters y for each group(Gibbons, 1971). To approximate the P-value of this statistic,the bootstrap samples must be drawn from the distribution thatsatisfies the relevant null hypothesis. This was accomplishedby resampling in three stages: (1) within each group, n fishwere sampled with replacement (2) ignoring originalgroupings, group labels were randomly assigned such thatthere were n fish in each group and (3) a smooth bootstrapsample of m fibres for each fish was generated. After 100bootstrap replications, the P-value was approximated by:

where #[D*>D] is the number of D*>D, and using the resultsD1*, ....., DR* from the R bootstrap samples (Davison andHinkley, 1997). Since the sample size of fish within a group issmall, 4–5 fish, the null hypothesis will probably not berejected if a significance level of 0.05 is required unless a largereal difference exists. Since significance level and power arerelated, given a fixed sample size (Zar, 1996), the criticalsignificance level was increased to 0.1 as suggested by Steeland Torrie (1980) to increase the power of the test.

This Kolmogorov–Smirnov-based bootstrap test is a globaltest, and a significant result strengthens the evidence that thedensities are different, but the test statistic alone provides fewclues to where the differences might occur. To supplement thistest, the two density curves were compared graphically. Firstf̂ D and f̂ T were plotted. If the null hypothesis was true, it wouldbe natural to ignore group labels and pool fibre diameters overall fish sampled in an age class when estimating the averagedensity (f̂ D+T) for an age class. To assess where the differencesbetween f̂ D and f̂ T can be attributed to a genuine difference instructure and where they can be attributed to random variation,a variability band for f̂ D+T was constructed. The region wheref̂ D and f̂ T lie outside this band suggests a major difference

1 + #[D*ùD]R + 1

P = , (3)

(2)Dmn = Smn(y) − Tmn(y) ,max

y

|||

|||

Muscle fibre diameter (µm)

Prob

abili

tyde

nsity

func

tion 0.01

0

A

0 50 100 150 200

B0.025

00 50 100 150 200

Fig. 2. A comparison of the histogram (A) and mean probabilitydensity function (B) for muscle fibre diameters from a triploidnormal-sex-ratio Atlantic salmon 775 days post-hatch.

1996

between the densities. Since this band provides a means ofevaluating the differences between f̂ D and f̂ T, it is referred toas a reference band.

The final stage of evaluating differences between groups wasto compare the values of specific percentiles of the estimatedfibre densities for each fish. The Wilcoxon two-sample non-parametric test was used to test whether the median value ofthe specified percentile was equivalent between groups.Specified percentiles were the fifth, tenth, fiftieth, ninety-fifthand ninety-ninth percentiles. Since we were not taking intoaccount the variance of the percentile estimate for each fish,and we were making multiple comparisons, the reported P-values should be interpreted with caution.

ResultsEmbryonic development

The unsegmented paraxial mesoderm started to segmentapproximately 17 days post-fertilization at 6 °C, producing

transient epithelial spheres (Fig. 3A) that subsequentlydifferentiated into mesenchymal derivatives. The somites weredifficult to count accurately with a binocular microscope untilthe following day. Mitotic bodies were relatively common inthe cells of the epithelial somites (arrows in Fig. 3A). Newsomites were added at the rate of approximately one every 6 huntil the full complement of 62–63 somites was formedapproximately 35 days post-fertilization (Fig. 4). ANCOVA ofsix families of NSR salmon revealed significant familydifferences in the rate of somite formation (F5,296=2.27;P=0.047), but no effects of ploidy.

An examination of serial sagittal sections indicated that thefirst myotubes formed adjacent to the notochord at the level ofthe horizontal septum (long arrows in Fig. 3B). Multi-nucleated myotubes formed in a gradient away from thehorizontal septum. Myoblasts apparently exited the cell cyclefollowing fusion since mitotic bodies were never observedwithin myotubes (Fig. 5A,B). A wave of myotube formation


Fig. 3. (A) A sagittal section of a normal-sex-ratio diploid salmon embryo with epithelialsomites (es) sampled 22 days post-fertilizationat the 28-somite stage and stained withhaematoxoylin–eosin. Somites 18–23, countingfrom the head, are illustrated. Arrows indicatecells actively involved in mitosis. Scale bar,25 µm. (B) Sagittal section of a normal-sex-ratiotriploid salmon embryo at the 46-somite stagestained with haematoxoylin–eosin. Long arrowsindicate mononuclear myotubes at the position ofthe horizontal septum. The short arrow indicatesa mitotic body in the sclerotome associated witha non-muscle cell. Note that some nuclei havethree nucleoli (tn). Scale bar, 25 µm.


progressed down the length of the trunk, starting in themost anterior somites at approximately the 30-somite stage(Fig. 6A). ANCOVA revealed small but statistically

significant effects of family (F5,66=9.54; P<0.001) and ploidy(F1,66=8.01; P=0.006) on myotube formation, with nosignificant two-way interaction (F5,66=1.64; P=0.16). Therelationship between the most posterior somite with myotubesand somite interval for two of the families is illustrated inFig. 6A. One somite interval is the mean time required for theformation of one somite pair. The rostral-to-caudal sequence

Age post-fertilization (days)

Num

ber

of s

omite

s

0

10

20

30

40

50

60

70

15 20 25 30 35 40

Fig. 4. Somite formation at 6 °C in full-sibling normal-sex-ratiodiploid (open symbols) and triploid (filled symbols) families ofAtlantic salmon. Only the data from families 1 and 5 are shown forclarity. Second-order linear regressions were fitted to the data.Family 1 (circles): for diploids (line omitted for clarity), N=−89.22+6.40t−0.049t2 (r2 adjusted=0.96; P=<0.001); for triploids(dashed line), N=−137.13+11.10t−0.156t2 (r2 adjusted=0.92;P=<0.001). Family 5 (triangles): for diploids (dotted line), N=−154.29+11.84t−0.161t2 (r2 adjusted=0.96; P=<0.001); fortriploids (solid line), N=−139.28+10.82t−0.144t2 (r2 adjusted=0.96;P=<0.001), where N is the number of somites and t is age in days.

Fig. 5. Sagittal sections of normal-sex-ratiosalmon embryos stained with haematoxoylin–eosin. (A) Diploid embryo sampled 25 dayspost-fertilization at the 42-somite stage. Somite18 is illustrated. Arrowheads indicate multi-nucleated myotubes. Note that the nuclei have amaximum of two nucleoli. Scale bar, 15 µm.(B) Triploid embryo sampled 42 days post-fertilization at the end of somitogenesis. Somites31 and 32 are illustrated. Arrows indicate cells inthe myosepta. Note that some of the nuclei havethree nucleoli (tn). Scale bar, 25 µm. ms,myoseptum.

1998

of myotube formation was somewhat more advanced withrespect to somite interval in the majority of triploid than ofdiploid families. The adjusted means for families 1–6 weresomite 38.3, 27.3, 24.3, 28.1, 32.9 and 29.1, respectively,whereas the adjusted means for ploidy were somite 28.06 fordiploid fish and 31.91 for triploid fish.

Myofibrils started to form at the 30- to 40-somite stage(Fig. 6B). The rostral-to-caudal sequence of myofibrilassembly is illustrated for two of the families in Fig. 6B.ANCOVA revealed significant differences for main effectsbetween families (F5,67=6.68; P<0.001) and with ploidy(F1,67=7.76; P=0.007), with no significant family-by-ploidyinteraction (F5,67=0.35; P=0.88). Myotubes containing

myofibrils are illustrated in Fig. 5B. Myofibril assemblycoincided with the appearance of cells closely associated withthe myosepta (arrows in Fig. 5B). Nuclei within myotubesfrom diploid fish contained one or two nucleoli (Fig. 5A),compared with one, two or three nucleoli in triploids(Fig. 5A,B). In the majority of families, myofibrillar assemblyhad reached more posterior somites at any given somiteinterval in triploid than diploid embryos. The adjusted meansfor families 1–6 were somite 31.5, 24.0, 19.8, 24.7, 29.4 and23.4, and the adjusted means for ploidy were somite 23.51 fordiploid fish and 24.45 for triploid fish.

Acetylcholinesterase (AChE) staining was concentrated atthe myosepta (Fig. 7A) and in the cell bodies and axons ofmotor neurones within the spinal cord (Fig. 7B). Therelationship between the most posterior somite with AChEstaining at the myosepta and somite interval for two of thefamilies is illustrated in Fig. 6C. ANCOVA revealedsignificant main effects of family (F5,137=17.81; P<0.001) andploidy (F1,137=8.36; P=0.004) on the progression of AChEstaining along the trunk with no significant family-by-ploidyinteraction (F5,137=0.91; P=0.48). AChE staining at themyosepta occurred slightly earlier with respect to somiteinterval in triploid than in diploid embryos. The combinedadjusted means for families 1–6 were somite 41.2, 39.2, 33.9,39.0, 47.5, and 43.4 and the adjusted means for ploidy weresomite 39.28 for diploid fish and 42.1 for triploid fish.

Organogenesis was investigated in relation to both somitestage and age for eight NSR (six from the 1995 series and twofrom the 1996 series) and two AF (1995 series) families oftriploid and diploid fish (Table 1). Although there were smallfamily differences in the somite stage at which somemorphological characters were first visible using DIC optics,no consistent differences between diploid and triploid fish


20 40 60 80 100 1200

10

20

30

40

50

60

70

Somite interval20 40 60 80 100 120

Som

ites

with

AC

hE s

tain

ing

0

10

20

30

40

50

60

70 C

20 40 60 80 100 120

Som

ites

with

myo

tube

s

0

10

20

30

40

50

60

70

B

A

Som

ites

with

myo

tube

s

Fig. 6. The relationship between the most posterior somite with thecharacter of interest, N, namely myotubes (A), myofibrils (B) andacetylcholinesterase (AChE) staining at the myosepta (C) and somiteinterval at 6 °C, in full-sibling normal-sex-ratio diploid (opensymbols, dotted lines) and triploid (filled symbols, solid lines)families of Atlantic salmon (Salmo salar L.) reared at 6 °C. Somiteswere numbered starting from the head. One somite interval (SI) isthe mean time taken for the formation of one somite pair, and its useallows the embryo to be staged beyond the period of somitogenesis.Only the data from family 1 (circles) and family 5 (triangles) areillustrated for clarity. First-order linear regressions were fitted to thedata. (A) N=myotubes. Family 1: diploids, N=−14.07+0.85SI(r2=0.97; P=<0.001); triploids, N=−15.11+0.96SI (r2=0.93;P=<0.001). Family 5: diploids, N=−11.94+0.74SI (r2=0.96;P=<0.001); triploids, N=−8.36+0.65SI (r2=0.98; P=<0.001). (B) N=myofibrils. Family 1: diploids, N=−8.39+0.58SI (r2=0.96;P=<0.001); triploids, N=−12.31+0.68SI (r2=0.93; P=<0.001).Family 5: diploids, N=−33.48+0.99SI (r2=0.97; P=<0.001); triploids,N=−30.64+0.96SI (r2=0.92; P=<0.001). (C) N=AChE. Family 1:diploids, N=−43.73+0.94SI (r2=0.93; P=<0.001); triploids, N=−46.32+1.00SI (r2=0.94; P=<0.001). Family 5: diploids, N=−10.04+0.62SI (r2=0.86; P=<0.01); triploids, N=−15.90+0.70SI(r2=0.88; P=<0.01).


could be detected with the frequency and density of samplingundertaken. We therefore conclude that the triploidisationprocedure produces at most only very minor changes in thetiming of organogenesis. Hatching started 81 days post-fertilization at 6 °C for both groups and continued for 10–12days.

Somatic growth

The relationship between fork length and age post-hatch wasdescribed by second-order polynomials and was notsignificantly different between diploids and triploids in NSR(F3,80=2.88; P=0.082) (Fig. 8A) and AF (F3,77=2.55; P=0.12)fish (Fig. 8B). However, diploids had a higher body mass fora given fork length than triploids for both NSR (F3,39=7.17;P<0.02) (Fig. 8C) and AF (F3,46=18.44; P<0.001) fish(Fig. 8D). For example, at 40 cm fork length in NSR groups,the mean body mass calculated from the regression equations

in Fig. 8C was 7.6 % greater for diploid (832.9 g) than fortriploid (747.2 g) fish. There was a low (<5 %) incidence ofskeletal deformities in triploid groups, with the mostcommon abnormalities being pronounced lower jaw growth,craniomegaly and twisted caudal vertebrae.

Muscle growth in freshwater stages

Muscle growth is a complex process involving therecruitment and hypertrophy of successive cohorts of musclefibres. In NSR diploids, the mean number of muscle fibres permyotome at the level of the pelvic fin insertions had increasedfrom 2786 in embryos 46 days post-fertilization to 5374following hatching (85 days post-fertilization) and had almostdoubled to 9852 by the time more than 95 % of the yolk hadbeen reabsorbed and the fish were ready to feed (Fig. 9). Athatching, the production of white muscle fibres was restrictedto germinal zones at the ventral and dorsal (arrowheads in

Fig. 7. Acetylcholinesterase staining in whole-mount Atlantic salmon (Salmo salar L.)embryos. (A) A 45-somite stage normal-sex-ratio triploid embryo 39 days post-fertilization.Note that the staining is largely confined tothe myosepta (arrows). Scale bar, 100 µm. (B) Motor neurone cell bodies in the spinal cordof a diploid Atlantic salmon embryo sampled 27days post-fertilization at the 45-somite stage.The position of the spinal cord is marked bylarge arrows. Small arrows indicate primarymotor neurone cell bodies, and pairs ofarrowheads illustrate the cell bodies ofsecondary motor neurones. Scale bar, 25 µm. a,anus; ff, primary fin-fold; g, gut; sc, spinal cord.

2000

Fig. 10A) apices of the myotomes. By first feeding, newlyrecruited very small-diameter white muscle fibres were evidentscattered throughout the myotome in both diploid (arrows inFig. 10B) and triploid (not illustrated) fish, consistent with theactivation of the satellite cell population. A general linear modelANCOVA was run for yolk-sac stages with age and ploidy asmain effects, fish length as covariate and fibre number anddiameter as dependent variables. For NSR fish, the number ofmuscle fibres per myotome increased with age (F3,36=12.02;P<0.001) and was significantly greater in diploids than triploids

(F1,36=5.85; P=0.022), with no significant age-by-ploidyinteraction. The differences in muscle fibre number wererelatively small for embryos, 7.0 % at 46 days post-fertilizationand 7.3 % at hatching (Fig. 9C). In contrast, there were 31 %more white muscle fibres in diploid than in triploid fish at firstfeeding (Fig. 9D). There was still a significant main effect ofploidy when the ANCOVA was re-run with only the embryosand newly hatched alevins (F1,25=13.03; P=0.001), withweighted mean numbers of fibres per myotome of 3894 fordiploid fish and 3599 for triploid fish, i.e. 7.6 % higher in


Table 1. Embryological development in Atlantic salmon (Salmo salar L.) reared at 6 °C

Time post-fertilization(days) Somite stage Developmental features

14 0 Embryos still in gastrulation stages15 0 Neural tube formation; 20–30 % epiboly17 6 50 % epiboly; germ ring and embryo distinct; some somites visible18 9–11 Formation of brain and eye vesicles; 60–90 % epiboly19 15–16 Development of eyes (visible laterally)20 18–21 Otic vesicle formed; 90–100 % epiboly21 22–24 Start of segmentation of hindbrain into five rhombomeres; development of branchial chamber; cardiac tube

in pericardial chamber22 27–28 Forebrain/midbrain boundary appearing; optic lens visible23 31–34 Gut and stomach primordia visible; myotomes acquiring chevron shape; midbrain/hindbrain boundary

apparent; one branchial arch visible; heart tube bending forwards24 36–37 Choroid fissure complete; lumen in stomach and gut; operculum visible25 38–40 Hindgut formation; tail detached from yolk at most caudal somites; otoliths present in otic vesicles26 41–44 Separation of ventral mid-body from yolk; heart tube moved to the left; slow heart beat apparent; posterior

end of pronephros level with end of gut; hindgut turned ventrally towards future anus27 45–47 Blood flow in dorsal aorta; two branchial arches; pectoral fin buds at somites 2–4

47–50 Weak sporadic contractions of the trunk28 49–52 Two-chambered heart with double beat and faster heart rate; movement in response to dechorionation; lens

separate from optic cup; rhombomere divisions starting to disappear; anus formed at level of somites 37–3829 53–55 Circulation visible along gut; rhombomere divisions almost disappeared; jaw developing; mesenchyme in

caudal fin-fold32 62–64 Rhythmic movements observed in situ; strong movements in response to dechorionation; olfactory organ

present; hatching glands apparent; dorsal and ventral fin-folds present; rhombomeres no longer distinct;three branchial arches; pectoral fin as high as it is wide

33 63–64 Thickening of fin-folds35 Rudimentary cloaca; liver primordium; vitelline vein functioning; cerebral circulation apparent36 Pigmentation over part circumference of eye; gill slit in anterior branchial chamber; heart valves present37 Four branchial arches; operculum covers base of first branchial arch; caudal end of notochord curves

upwards39 Pigmentation of whole of circumference of eye; haemoglobin present; eyes visible through chorion;

intestinal arteries and gut capillaries functional42 Approximately two-thirds of yolk sac vascularised; liver lobulated and vascularised45 Superior semicircular canal complete; lateral head movements – lower jaw free of yolk sac49 Complete vascularisation of yolk; circulation in dorsal aorta extends to tail; early caudal fin rays

apparent; circulation in branchial arches and olfactory pits; all semicircular canals complete55 Branchial arch blood vessels complete; blood flow in network of fine vessels across brain; fin-fold shaped

inwards at anus; formation of dorsal and anal fin rays; skin pigmentation apparent on dorsal surface ofhead and trunk

70 Spontaneous flapping movement of pectoral fins; jaw movements; formation of gill filaments77 Elongation of head and snout; pelvic fin buds apparent; operculum covers 3–4 of the gill arches; elongation

of yolk sac; skin pigmentation visible through chorion; pectoral fin rays present81 Hatching begins

Eight families of normal-sex-ratio diploid and triploid fish and two families of all-female fish were examined.


diploids. In contrast, for AF fish, the main effect of ploidy onthe number of muscle fibres per myotome was not significantlydifferent for similar stages (Fig. 9A–C) (F1,26=3.63; P=0.068).

The mean diameter (µm) of white muscle fibres wassignificantly greater in triploids than in diploids for both NSR(F1,29=9.81; P=0.004) and AF (F1,26=76.9; P<0.001) fish, with

Table 2. The number of white muscle fibres per myotome in diploid and triploid Atlantic salmon (Salmo salar L) at various stagesof the life cycle

Normal-sex-ratio All-female fish

Stage Diploid Triploid Diploid Triploid

S1 parr 83 509±6758 59 529±3343 62 084±4777 58 789±4285(4) (5) (5) (5)

[8.2 cm] [8.2 cm] [7 cm] [7 cm]

Seawater- 292 700±30 996 154 340±8039 195 580±19 628 160 120±7997adapted smolt (5) (5) (5) (5)

[18 cm] [18 cm] [14.6 cm] [17.2 cm]

One sea-winter 583 554±15 950 369 954±22 791 553 789±39 395 383 961±25 320(6) (4) (5) (8)

[38–42 cm] [38–42 cm] [38–42 cm] [38–42 cm]

Values are presented as means ± S.E.M. for (N) fish.All measurements were made at the level of the pelvic fin insertions.The mean or range of fork length of each stage is shown in square brackets.


Fork

leng

th (

cm)

0

10

20

30

40

50

NSR

0

200

400

600

800

1000

All-female

Fork length (cm)

Bod

y m

ass

(g)

0

200

400

600

800

1000

NSR


Fork

leng

th (

cm)

0

10

20

30

40

50

All-female

A B

C D

0 200 400 600 800 1000 0 200 400 600 800 1000

15 20 25 30 35 40 45

Bod

y m

ass

(g)

Fork length (cm)15 20 25 30 35 40 45

Fig. 8. The growth of normal-sex-ratio (NSR) (A,C) and all-female (B,D) Atlantic salmon (Salmo salar L.) at ambient temperature. The arrowindicates transfer of smolts from fresh water to sea water. Open symbols and dotted lines are for diploids, and filled symbols and solid lines arefor triploids. The data were fitted with a series of quadratic equations with the following formulae: (A) for normal-sex-ratio diploids,FL=2.96+0.017t+2.95×10−5t2 (r2 adjusted=0.97; P<0.001) (where FL is fork length and t is days post-hatch); for normal-sex-ratio triploids,FL=2.60+0.017t+3.47×10−5t2 (r2 adjusted=0.97; P<0.001); (B) for all female diploids, FL=2.93+0.0017t+5.07×10−5t2 (r2 adjusted=0.97;P<0.001), for all-female triploids, FL=2.23+0.013t+3.83t2 (r2 adjusted=0.97; P<0.001); (C) for normal-sex-ratio diploids,M=390.9−42.15FL+1.33FL2 (r2 adjusted=0.99; P<0.001) (where M is body mass in g); for normal-sex-ratio triploids,M=331.2−35.63FL+1.15FL2 (r2 adjusted=0.99; P<0.001); (D) for all-female diploids, M=208.8−26.69FL+1.02FL2 (r2 adjusted=0.99;P<0.001); and for all-female triploids, M=149.7−20.03FL+0.84FL2 (r2 adjusted=0.99; P<0.001).

2002

adjusted mean values of 7.76 µm (triploid) and 6.44 µm(diploid) and 5.37 µm (triploid) and 4.50 µm (diploid),respectively (Fig. 11) (all stages). The number of white musclefibres increased approximately 30-fold in NSR salmon betweenthe alevin and smolt stages (Table 2).

Satellite cell densities in smolts

The number of satellite cells was estimated in 7 µm thickfrozen sections using antibodies against c-met and myf-6. Thec-met receptor tyrosine kinase is located on the cell surface,whereas myf-6 is expressed in the nucleus. The dimensions ofsatellite cells were investigated using transmission electronmicroscopy in sagittal sections of white muscle from diploidfish. Satellite cells are spindle-shaped cells approximately2–5 µm in diameter and 7–14 µm long with a separatemembrane from the adjacent muscle fibres, containing aprominent nucleus (Fig. 12A,B). At first feeding, the satellitecells were relatively abundant, and most were located betweenthe muscle fibres (Fig. 12A). In sagittal section, the nucleusoccupied approximately 0.40±0.14 (mean ± S.D.) of the lengthof the cell (measurements from 12 cells). Ultrathin (60 nm)transverse sections at the level of the satellite cell nucleusrevealed a relatively sparse cytoplasm with occasional roughendoplasmic reticulum (Fig. 12A).

Frozen sections were stained with antibodies to c-met(Fig. 13A,B) and myf-6 (Fig. 13C). In estimating the numberof nuclei in c-met-positive cells, a correction factor wasapplied to account for the greater probability of encounteringthe cytoplasm than the nucleus in transverse sections ofsatellite cells. Control sections with either the primary orsecondary antibodies omitted were completely unstained. Thedensity of c-met-positive cells was 24 % higher in diploid thanin triploid smolts (F1,8=5.98; P=0.040) and the density of myf-6-positive cells was 34 % higher (F1,8=8.10; P=0.022)

(Table 3). The ratio of nuclei in c-met-positive cells to totalmyonuclei was approximately 0.175, irrespective of ploidy. Insmolts, 22 cm mean length, there were 23 % more myofibrenuclei of all classes per unit volume in diploids than in triploids(F1,8=10.4; P=0.012) (Table 3).

Muscle growth in seawater stages

The total cross-sectional area of white muscle in NSRdiploids increased more than 3100-fold between the alevinstage and the end of first year of seawater growth to 42 cm forklength (Fig. 14A,B). In smolts, 18 cm fork length, the totalcross-sectional area of white muscle was similar in NSRdiploids (511 mm2) and NSR triploids (478 mm2) (Fig. 14A).However, the hypothesis that the total cross-sectional area ofwhite muscle fibres was the same in diploid and triploid fishwas rejected for both NSR (F3,78=7.48; P<0.001) (Fig. 14A)and AF (F3,76=4.18; P=0.017) fish (Fig. 14B) because of


Num

ber

of f

ibre

s

0

1000

2000

3000

4000

5000

NSR All-female

46-day embryo

0

1000

2000

3000

4000

5000 57-day embryo

NSR All-female

Num

ber

of f

ibre

s

0

2000

4000

6000

8000

10000

12000

NSR

Hatch

0

2000

4000

6000

8000

10000

12000

NSR

First feeding

A B

C D

All-female

Fig. 9. The number of fast musclefibres per myotome in the embryonicand yolk-sac stages of normal-sex-ratio (NSR) and all-female diploid(open columns) and triploid (filledcolumns) Atlantic salmon reared atambient temperatures. (A) 46-dayembryos, (B) 57-day embryos, (C) 1-day-old alevins (Hatch) and (D) fishat first feeding (all-female fish notsampled). Values are means + S.D. forfive or six fish per sample point.

Table 3. Density of c-met- and myf-6-positive nuclei and totalmyofibre nuclei in the fast muscle of normal-sex-ratio diploid

and triploid salmon smolts

Number of nuclei per mm3

muscle cross-sectional area

Antibody/or other stain Diploids Triploids

c-met 12 197±2423 9279±1083myf-6 10 200±2329 6700±1486Meyer’s haematoxylin 69 087±6943 53 544±8372

(total myofibre nuclei)

A correction was made for section thickness and the relativeprobability of encountering the cytoplasm or nucleus of amononuclear cell stained with c-met (see text for details).

Results are presented as ±S.D. for five fish per group.


differences in the larger fish. This may in part reflectdifferences in body shape between diploids and triploids aswell as a small difference in body mass for a given fork length.In 42 cm fish, the cross-sectional area of white muscle was2995 mm2 in NSR diploids compared with 2533 mm2 in NSRtriploids, i.e. 14.8 % greater. At this body length, the cross-sectional area of white muscle was approximately 4.8 %greater in AF diploids than in triploids.

The number of white muscle fibres increased approximately3.5-fold during the first year of seawater growth. Musclerecruitment was investigated by fitting a second-orderpolynomial to the estimates of the number of white musclefibres per myotome in the different age classes (Fig. 15A,B).ANCOVA with fish length as covariate showed significantmain effects of age class and ploidy on the number of whitemuscle fibres per myotome for both NSR and AF populations(P<0.001). Tukey post-hoc tests revealed significant

differences in muscle fibre number between diploids andtriploids for all the seawater stages of the life-cycle (P<0.05).After 800 days, the mean number of fast muscle fibrescalculated from these relationships was 540 051 for NSRdiploids, 363 596 for NSR triploids, 525 361 for AF diploidsand 369 891 for AF triploids. The mean rates of muscle fibrerecruitment were found by differentiating the second-orderregressions in Fig. 15A,B. Rates of recruitment 800 days post-hatch were 931 and 659 fibres day−1 for diploids and triploids,respectively, in the NSR groups (Fig. 15C) and 1162 and608 fibres day−1 in diploids and triploids, respectively, in theAF fish (Fig. 15D).

The relationship between the number of muscle fibres permyotome at the level of the pelvic fin insertions and fork lengthis illustrated in Fig. 16A,B. For 42 cm fork-length fish, therewere approximately one-third more muscle fibres per myotomein diploid than in triploid groups, 649, 878 and 413 619,

Fig. 10. Muscle cellularity in Atlantic salmon(Salmo salar L.). Transverse wax sectionsstained with haematoxylin–eosin. (A) A normal-sex-ratio diploid fish at hatching. The regionillustrated is the dorsal epaxial muscle at thelevel of the pelvic fin insertions. Arrowheadsshow the germinal zones of new muscle fibreproduction. Scale bar, 20 µm. (B) A normal-sex-ratio diploid fish at first feeding. The arrowsindicate new muscle fibres formed from theactivation of the satellite cell population.Arrowheads indicate nuclei. Scale bar, 10 µm.r, red muscle fibres; w, white muscle fibres; mc,mucus cell; sk, skin; sc, spinal cord.

2004

respectively, for all-female fish. The major differences inmuscle cellularity arising from the higher rates of fibrerecruitment and lower rates of hypertrophic growth in diploidthan in triploid groups are illustrated in Fig. 16C,D by plottingthe number of fibres against the total cross-sectional area ofwhite muscle. In general, for the seawater stages of the lifecycle, there were approximately one-third more muscle fibresper unit of muscle cross-sectional area in diploid than intriploid fish in both groups.

The distribution of muscle fibre size: estimating hypertrophicgrowth

The estimated probability density function (PDF) for eachfish in the age classes sampled and the mean PDF for all fishare shown for the NSR (Fig. 17) and AF groups (Fig. 18).

Variation between individual fish generally increased with age(Figs 17, 18). At first feeding in NSR diploid salmon, the peakPDF corresponded to a muscle fibre diameter of approximately20 µm. For individual NSR diploids, several peaks in the PDFwere apparent at 315 days, 470 days and 620 days post-hatch,indicating cycles of new muscle fibre recruitment (Fig. 17).Variability bands for the mean PDF for these ages (Fig. 19)provided no evidence for a multimodal distribution. Althoughthe PDF was broader in these age classes, the peak stillcorresponded to a diameter of 20 µm. The maximum musclefibre diameter in NSR fish increased from 45 µm at first feedingto 120 µm at 315 days post-hatch and was approximately200 µm in the 775 day and 839 day age classes.

Distinct bimodal distributions of muscle fibre diameter wereapparent and supported by the variance bands (Figs 17–20) in


Fibr

e di

amet

er (

µm)

0

2

4

6

8

10 46-day embryo

NSR All-female0

2

4

6

8

10

NSR All-female

57-day embryo

Fibr

e di

amet

er (

µm)

0

2

4

6

8

10

NSR All-female

Hatch

0

4

8

12

16

20

NSR

First feeding

A B

C DFig. 11. The mean diameter of fast musclefibres in the embryonic and yolk-sac stages ofnormal-sex-ratio (NSR) and all-female diploid(open columns) and triploid (filled columns)Atlantic salmon reared at ambienttemperatures. (A) 46-day embryos, (B) 57-dayembryos, (C) 1-day-old alevins (Hatch) and(D) fish at first feeding (all-female fish notsampled). Values are means + S.D. for five orsix fish per sample point.

Fig. 12. Electron micrographs of sagittallysectioned white muscle fibres from diploidnormal-sex-ratio salmon at first feedingillustrating (A,B) satellite cells and (C)myofibre nuclei. The arrowhead indicatesrough endoplasmic reticulum and the doublearrowheads indicate the muscle fibresarcolemma. mt, mitochondrion; mf,myofibril; n, myofibre nucleus; s, satellitecell; sn, satellite cell nucleus; z, Z-line. Thescale bars represent 1 µm in A, 2 µm in Band 3 µm in C.


both the NSR and AF groups sampled 775 days and 839 dayspost-hatch, reflecting seasonal cycles of seawater growth. The

density of muscle fibres was compared using a non-parametricbootstrap test and reference band, which represents the nullhypothesis that there was no difference between the groups.Regions of the curves where the density estimates (broken line,diploids; solid line, triploids) were out of the range of the 100bootstrapped density estimates indicated a difference betweendiploid and triploid fish (Figs 21, 22). Test results betweengroups with age class are shown in Table 4. For NSR fish, theprobability density curves were not significantly differentbetween diploids and triploids at first feeding (101 days post-hatch) (P=0.32), but there was evidence for significantdifferences for all other age classes (P=0.07) (Table 4).

For AF fish, the density curves were significant for 315 days,775 days and 839 days (P=0.07), but not significant for 470days and 620 days (Table 4). For both these age classes, thereis one fish whose estimated PDF is distinctively different fromthat of the other fishes in the group. When theKolmogorov–Smirnov test statistics were significant, thereference bands were used to help determine where thedifferences occurred. Reference bands for NSR fish at 315 daysand 470 days indicate that the diploids had a higher percentageof smaller fibres and the triploids had a thicker right-hand tail.For 620, 775 and 839 days, the reference bands indicated majordifferences in the middle ranges of the distributions, withtriploids having a thicker right-hand tail. Similar patternsexisted for the AF fish at 315 days, 775 days and 839 days, butthe evidence was not as strong. The fifth, tenth, fiftieth, ninety-fifth and ninety-ninth percentiles were calculated for the meanprobability density function of each age class. Since the datawere found to have non-constant variance and non-normalerrors, a non-parametric Wilcoxon test was used to give anindication of differences between diploids and triploids in

Fig. 13. Immunoperoxidase staining of frozen sections of whitemuscle from normal-sex-ratio salmon smolts: red reaction productindicates cells positive for the antigen (arrows) (A,B) Sectionsstained with anti-c-met. (A) Muscle section from a triploid fishwithout counterstain. Scale bar, 50 µm. (B) Section from a diploidfish counterstained with Mayer’s haematoxylin (blue staining). Pairsof arrowheads indicate double-labelled cells. Scale bar, 25 µm.(C) Section from a diploid fish stained for myf-6 withoutcounterstain. Scale bar, 25 µm.

Table 4. Non-parametric bootstrap test for fT=fD using theKolmogorov–Smirnov statistic, where fD and fT represent themean probability density functions for muscle fibre diameters

of diploid and triploid Atlantic salmon, respectively

Age post-hatch(days) Approximate P-value

Normal-sex-ratio fish101 0.32315 0.05470 0.07620 0775 0.04839 0.07

All-female fish315 0.06470 0.32620 0.18775 0.03839 0.07

P-values of approximately 0.1 were considered to indicatesignificant differences in the overall distribution.

2006

cases where the Kolmogorov–Smirnov test statistic wassignificant (Tables 5, 6). For NSR fish (Table 5), the fiftieth,ninety-fifth and ninety-ninth percentiles of the mean PDF forthe triploids were consistently higher than the percentiles fordiploids except at 839 days, where the differences were notsignificant. This relationship appears to hold at 315 days and775 days for the AF fish (Table 6), but the evidence is not asstrong. The fiftieth, ninety-fifth and ninety-ninth percentiles ofthe mean PDF in AF fish at 839 days were similar for diploidsand triploids, as for the NSR groups.

DiscussionMyogenesis is a complex process involving the proliferation

of mesodermal stem cells, commitment to a specific muscle-fibre lineage, exit from the cell cycle and terminaldifferentiation (Stockdale, 1992). The majority of embryonicmuscle fibres in fish are formed by the fusion of severalmyoblasts to form multi-nucleated myotubes that subsequentlyexpress contractile proteins (Waterman, 1969; Van

Raamsdonk et al., 1974; Johnston et al., 1995). Increasedheterozygosity and lower numbers of cells would be expectedto decrease, and larger cell size to increase, rates ofdevelopment in triploid compared with diploid fish (Leary etal., 1985). In the present study, we found evidence thatmyogenesis and neuromuscular development occurred atslightly earlier somite stages in triploids than in diploids,although the differences were minor and smaller than thevariation among different families. A small decrease in thetime between fertilization and hatching was noted in triploidrainbow trout (Oncorhyncus mykiss) (Happe et al., 1988;Quillet et al., 1988). However, salmonids can hatch over arange of developmental stages (Killeen et al., 1999), and wefound no evidence that triploidisation significantly affected thetiming of organ formation relative to somite stage (Table 1).

Studies in zebrafish have shown that the myoblasts that giverise to embryonic red muscle fibres originate from rows oflarge cuboidal cells adjacent to the notochord that can beidentified by their expression of MyoD and snail 1 prior tosomite formation (Weinberg et al., 1996). These adaxial cellscommit to the red muscle lineage under the influence of theglycoprotein Sonic hedgehog, secreted from the notochord,and migrate through the somite to a superficial position justunderneath the skin (Devoto et al., 1996; Blagden et al., 1997).A subset of these cells elongate to form mononuclear myotubes


Table 5. Comparison of the percentiles for the meanprobability density functions of muscle fibre diameter in

normal-sex-ratio diploid (fD) and triploid (fT) Atlantic salmon

Agepost-hatch Wilcoxon test Percentile Percentile(days) Percentile P-value for f D for f T

101 0.05 1.00 8.28 7.960.10 1.00 10.01 9.780.50 0.69 19.00 20.060.95 0.31 34.33 36.850.99 0.42 40.73 44.16

315 0.05 0.01 11.77 13.420.10 0.01 15.01 17.580.50 0.01 42.58 47.690.95 0.01 103.58 118.030.99 0.01 128.12 146.09

470 0.05 0.03 10.06 15.010.10 0.06 13.50 19.760.50 0.06 43.78 54.520.95 0.03 112.73 135.020.99 0.02 146.35 167.43

620 0.05 0.69 11.20 9.250.10 0.55 14.46 12.310.50 0.01 37.89 50.850.95 0.01 88.75 114.080.99 0.01 111.64 147.42

775 0.05 1.00 10.50 10.090.10 1.00 13.38 12.900.50 0.03 71.66 86.610.95 0.06 141.46 167.620.99 0.11 175.53 206.56

839 0.05 0.69 10.75 9.510.10 0.89 22.51 15.010.50 0.11 71.75 82.280.95 0.11 136.84 155.120.99 0.11 169.56 190.84

Table 6. Comparison of the percentiles for the meanprobability density functions of muscle fibre diameter in all-

female diploid (fD) and triploid (fT) Atlantic salmon

Agepost-hatch Wilcoxon test Percentile Percentile(days) Percentile P-value for f D for f T

315 0.05 0.11 11.53 13.040.10 0.06 14.33 17.630.50 0.03 39.23 47.580.95 0.20 104.56 115.830.99 0.20 129.51 143.50

470 0.05 0.15 12.24 13.470.10 0.15 15.61 16.940.50 0.15 41.81 46.960.95 0.22 133.72 120.300.99 0.22 208.26 143.20

620 0.05 0.84 10.03 9.300.10 0.84 13.24 11.990.50 0.55 39.33 42.480.95 0.22 96.07 103.720.99 0.31 129.49 130.89

775 0.05 0.84 11.28 10.910.10 0.84 14.66 13.850.50 0.03 64.50 76.530.95 0.06 128.50 145.660.99 0.06 160.15 179.94

839 0.05 0.49 10.83 12.300.10 0.49 18.88 24.780.50 0.34 73.68 81.300.95 0.11 133.90 151.180.99 0.11 165.19 186.93


and migrate through the myotome to form the muscle pioneerfibres (Devoto et al., 1996) that are thought to have a role in

guiding the growth cones of the primary motor neurones(Kimmel et al., 1991). Similar mononuclear myotubes were

Tota

l cro

ss-s

ectio

nal a

rea

(mm

2 )

0

1000

2000

3000

NSR

A

Fork length (cm)

0

1000

2000

3000

All-female

B

0 10 20 30 40 50 0 10 20 30 40 50

Fig. 14. The relationship between the total cross-sectional area (CA) of white muscle fibres in myotomes adjacent to the pelvic fin insertionsand fork length (FL) for (A) normal-sex-ratio (NSR) and (B) all-female Atlantic salmon (Salmo salar L.). Diploids, open symbols; triploids,filled symbols. The lines (dotted for diploids and solid for triploid fish) were fitted with a series of second-order linear regression equations withthe following formulae: for normal-sex-ratio diploids, CA=−12.89−2.74FL+1.77FL2 (r2 adjusted=0.95; P<0.001); for normal-sex-ratiotriploids, CA=9.86+0.43FL+1.42FL2 (r2 adjusted=0.97; P<0.001); for all-female diploids, CA=16.62−0.83FL+1.58FL2 (r2 adjusted=0.98;P<0.001); for all-female triploids, CA=20.70−1.96FL+1.46FL2 (r2 adjusted=0.97; P<0.001).

Num

ber

of m

uscl

e fi

bres

0

6×105

NSR All female

All-female


Num

ber

of f

ibre

s re

crui

ted

per

day

0

200

400

600

800

1000

1200

NSR

A B

C D

4×105

2×105

0 200 400 600 800 1000

0 200 400 600 800

Num

ber

of m

uscl

e fi

bres

0

6×105

4×105

2×105

0 200 400 600 800 1000

Num

ber

of f

ibre

s re

crui

ted

per

day

0

200

400

600

800

1000

1200

Age post-hatch (days)0 200 400 600 800

Fig. 15. (A,B) The relationship between the number of white muscle fibres in myotomes adjacent to the pelvic fin insertions and age post-hatch(days) for (A) normal-sex-ratio (NSR) and (B) all-female Atlantic salmon (Salmo salar L.). Diploids, open symbols; triploids, filled symbols.The lines (dotted for diploids and solid for triploid fish) were fitted with a series of second-order linear regression equations with the followingformulae: for normal-sex-ratio diploids, n=−5869.5+432.8t+0.31t2 (r2 adjusted=0.92; P<0.001) (where n is the number of muscle fibres); fornormal-sex-ratio triploids, n=−5564.3+339.9t+0.15t2 (r2 adjusted=0.89; P<0.001); for all-female diploids, n=17 481.13+52.4t+0.73t2 (r2

adjusted=0.90; P<0.001); for all-female triploids, n=−1013.2+319.6t+0.18t2 (r2 adjusted=0.87; P<0.001). (C,D) Muscle fibre recruitment in (C)normal-sex-ratio and (D) all-female Atlantic salmon (Salmo salar L.). The data were obtained by differentiating the equations in A and B usingthe Mathematica computing environment. The equations obtained were as follows: for normal-sex-ratio diploids (dotted line), r=439.8+0.614t;for normal-sex-ratio triploids (solid line), r=308.6+0.437t; for all-female diploids (dotted line), r=90.85+1.339t ; for all-female triploids (solidline), r=319.6+0.360t, where r is the rate of fibre recruitment (fibres per day).

2008

observed at the horizontal septum in Atlantic salmon (Fig. 3B).In zebrafish, the white muscle fibres were shown to originatefrom a distinct population of myoblasts in the lateral segmentalplate mesoderm (Devoto et al., 1996) that do not require Sonichedgehog for commitment (Blagden et al., 1997). It has beenreported that myotomes of rainbow trout contain only whitemuscle fibres at the yolk-sac stage, with the red muscle layerconsisting of undifferentiated myoblasts (Nag and Nursall,1972). In contrast, a single superficial layer of red muscle fibrescontaining myofibrils is already present in Atlantic salmonalevins at hatching (Fig. 10A). In their natal streams, the yolk-sac stages of both salmon and trout fry remain buried in thegravel and are not free-swimming.

Undifferentiated myoblasts which may contribute to thesatellite cell population were observed on the surface of theembryonic muscle fibres as early as 42 days post-fertilizationin Atlantic salmon (Fig. 5B). It has been suggested, on thebasis of ultrastructural observations on two species offreshwater fish, that satellite cells originate from the adjacentmesenchymal tissue lining and migrate into the muscle tissue

via the myosepta (Stoiber and Sanger, 1996). Although weobserved cells in the myosepta at approximately 42 days post-fertilization (Fig. 5B), these could just as conceivably havebeen cells associated with the synthesis of structural elementssuch as collagen fibres.

Foetal and neonatal myoblasts that are involved in theformation of primary and secondary muscle fibres,respectively, have been identified in mammals (Wilson et al.,1992; Stockdale, 1992). The full complement of muscle fibresis present at approximately the time of birth, and subsequentgrowth involves fibre hypertrophy only (Rowe andGoldspink, 1969). In the muscles of adult mammals, anotherpopulation of myoblasts can be identified that are referred toas satellite cells because of their anatomical position beneaththe basal lamina of muscle fibres (Mauro, 1961). Satellitecells represent approximately 24 % of the total myonuclei inthe biceps brachii of rapidly growing newborn mice (Brownand Stickland, 1993). The proportion of satellite cells to totalmyofibre nuclei falls dramatically with age in mice, reachingless than 1 % at 24 months (Gibson and Schultz, 1983). In


All-female

Total cross-sectional area (mm2)

NSR All-female

Fork length (cm)

Num

ber

of m

uscl

e fi

bres

0

6×105 NSRA B

C D

4×105

2×105

0 10 20 30 40 50Fork length (cm)

0

6×105

4×105

2×105

0 10 20 30 40 50

Num

ber

of m

uscl

e fi

bres

0

6×105

4×105

2×105

0 500 1000 1500 2000 2500 3000 35000

6×105

4×105

2×105

Total cross-sectional area (mm2)0 500 1000 1500 2000 2500 3000 3500

Fig. 16. (A,B) The relationship between the number of white muscle fibres in myotomes adjacent to the pelvic fin insertions (n) and fork length(FL) for (A) normal-sex-ratio (NSR) and (B) all-female Atlantic salmon (Salmo salar L.). Diploids, open symbols; triploids, filled symbols.The lines (dotted for diploids and solid for triploid fish) were fitted with a series of first-order linear regression equations with the followingformulae: for normal-sex-ratio diploids, n=−34 373.4+16291.0FL (r2 adjusted=0.94; P<0.001); for normal-sex-ratio triploidsn=−12 272.2+10 140.3FL (r2 adjusted=0.94; P<0.001); for all-female diploids, n=−34 373.4+16 291.0FL (r2 adjusted=0.90; P<0.001); for all-female triploids, n=−12 272.2+10 140.3FL (r2 adjusted=0.87; P<0.001). (C,D) The relationship between the number and the total cross-sectional area (CA) of white muscle fibres in myotomes adjacent to the pelvic fin for (C) normal-sex-ratio and (D) all-female Atlantic salmon.Diploids, open symbols; triploids, filled symbols. The lines (dotted for diploids and solid for triploid fish) were fitted with a series of third-orderlinear regression equations with the following formulae: for normal-sex-ratio diploids, n=20449.5+564.9CA−0.22CA2+ 3.29CA3 (r2

adjusted=0.91; P=<0.001); for normal-sex-ratio triploids, n=9586.6+448.5CA−0.21CA2+3.66CA3 (r2 adjusted=0.90; P=<0.001); for all-femalediploids, n=16620.9+552.8CA−0.21CA2+3.0CA3 (r2 adjusted=0.92; P=<0.001); for all-female triploids, n=10816.3+545.7CA−0.34CA2+7.83CA3 (r2 adjusted=0.90; P=<0.001).


adult mammals, satellite cells are usually quiescent,becoming activated and entering the cell cycle in response tomuscle damage, which indicates that they have a role in repairand regeneration (Schultz and McCormick, 1994). In mice,approximately 80 % of satellite cells divide with a cell cycletime of 32 h (Schultz, 1996). The remaining 20 % of satellitecells cycle more slowly and are thought to represent a reservepopulation of muscle stem cells. Muscle stem cells arethought to undergo an asymmetric division resulting in adaughter stem cell and an activated stem cell that can undergoa limited number of further divisions prior to terminaldifferentiation (Schultz, 1996).

Genetic analyses of knock-out mice have shown that the fourmembers of the MyoD gene family (MyoD, myogenin, myf-5and myf-6) play distinct but overlapping roles in myogenic

determination and differentiation (for reviews, see Olson andKlein, 1994; Yun and Wold, 1996). Activated satellite cellsaccumulate members of the MyoD and myocyte enhancerbinding factor-2 (MEF2) gene families of muscle transcriptionfactors, whereas myogenic regulatory factor (MRF) proteinsare not expressed in quiescent satellite cells (Grounds et al.,1992; Smith et al., 1994; Yablonka-Reuveni and Rivera, 1994;Cornelison and Wold, 1997). myf-5 and MyoD appear atrelatively early stages of the myogenic pathway in mammals,whereas myogenin and myf-6 are expressed at later stages(Smith et al., 1994). A proportion of MRF-positive cells alsoexpress m-cadherin (Cornelison and Wold, 1997) and Bcl-2, aprotein involved in the delay or suppression of apoptosis(Dominov et al., 1998). Cell cycle withdrawal and myogenicdifferentiation are also associated with the up-regulation of

101 days post-hatch, January 1995

315 days post-hatch, November 1995

470 days post-hatch, April 1996Pr

obab

ility

dens

ityfu

nctio

n

620 days post-hatch, September 1996

775 days post-hatch, March 1997

839 days post-hatch, May 1997


A

B

C

D

E

F

Diploids Triploids

Diploids Triploids

Diploids Triploids

Diploids Triploids

Diploids Triploids

Diploids Triploids

0.06

0

0.020

0

0.020

0

0.03

0

0.03

0

0.03

0

0.06

0

0.020

0

0.020

0

0.03

0

0.03

0

0.03

0

0 10 20 30 40 50 0 10 20 30 40 50

0 50 100 150 200 0 50 100 150 200

0 50 100 150 200 0 50 100 150 200

0 50 100 150 200 0 50 100 150 200

0 50 100 150 200 0 50 100 150 200

0 50 100 150 0 50 100 150

Fig. 17. Probability density functions ofmuscle fibre diameter in normal-sex-ratioAtlantic salmon 101–839 days post-hatch.Dotted lines illustrate the functions forindividual fish and the solid line is thecalculated mean value.

2010

expression of the general cyclin-dependent kinase inhibitorp21 (Walsh and Perlman, 1997). Studies on muscle satellitecells have been handicapped by the absence of suitablemolecular markers. Recently, Cornelison and Wold (1997)have shown that quiescent and activated mouse satellite cellsexpress the c-met receptor tyrosine kinase, which was notdetected on muscle-derived fibroblasts or other mononuclearcells. c-met is a membrane-bound disulphide-linkedheterodimer with an intracellular tyrosine kinase domain(Gonzatti-Haces et al., 1988) activated by the ligandhepatocyte growth factor/scatter factor (HGF/SF) (Matsumotoand Nakamura, 1996). HGF/SC is present in adult mammalianmuscle and is capable of activating satellite cells in tissueculture (Tatsumi et al., 1998).

In Atlantic salmon, c-met is present in mononuclear cellsthat also express proliferating cell nuclear antigen, an auxillaryprotein to DNA polymerase (results not illustrated), indicatingthat they also represent an actively dividing population of cells.The majority of c-met-containing cells (84 % in diploids and72 % in triploids) also expressed myf-6, which is consistent

with them representing a subset of proliferating cellscommitted to differentiation. On the basis of these markers,putative satellite cell nuclei in salmon smolts representapproximately 17.5 % of the total myofibre nuclei (Table 3),reflecting the rapid rate of recruitment and hypertophy ofmuscle fibres during the early saltwater stages. c-mettransduces migratory, mitogenic and morphogenic signals in anumber of non-muscle tissues during development and wound-healing as well as in satellite cells (Matsumoto and Kamura,1996). Before c-met can be definitively accepted as amolecular marker of satellite cells in salmon, further studies onits localisation at an ultrastructural level are required todetermine whether it can be expressed in any other cell types.We estimated that NSR diploids were recruiting more than1000 fibres myotome−1 day−1 during the second summer afterseawater transfer (Fig. 15A,C). Using ultrastructural criteria,Koumans et al. (1991) determined that, in the common carp(Cyprinus carpio), the percentage of satellite cell nuclei wasapproximately 8 % in rapidly growing juveniles approximately5 cm in standard length. The proportion of satellite cell nuclei


Prob

abili

tyde

nsity

func

tion



470 days post-hatch, April 1996


A

B

C

Diploids

Diploids Triploids

Diploids Triploids

0.03

0

0.03

0

0.03

0

Triploids

0 50 100 150

0.03

00 50 100 150

0 50 100 150 200 250 300 0 50 100 150 200 250 300

0 50 100 150 200 0 50 100 150 2000

0.03

0

0.03



D

E

Diploids

Diploids Triploids

0.03

0

0.03

0

Triploids0.03

0

0

0.03

0 50 100 150 200 250 0 50 100 150 200 250

0 100 200 300 0 100 200 300

Fig. 18. Probability density functions ofmuscle fibre diameter in all-femaleAtlantic salmon 315–839 days post-hatch. Dotted lines illustrate thefunctions for individual fish and thesolid line is the calculated mean value.


fell to less than 2 % of the total in fish 15–20 cm long(Koumans et al., 1991), a value significantly lower than that inAtlantic salmon, which have significantly faster growth ratesat this length.

The hypothesis that diploid fish would have a higher densityof satellite cells, leading to correspondingly faster rates of fibrerecruitment, was sustained. However, muscle fibres in triploidfish showed greater rates of hypertrophic growth than indiploids, although the patterns observed were complex andage-dependent. In the present study, we have applied non-parametric statistical techniques to compare the distribution offibre size in diploid and triploid groups to provide a more

quantitative description of the growth history of the successivecohorts of muscle fibres recruited. The majority of previousstudies of muscle growth in fish have been essentiallydescriptive, involving the measurement of the diameters orcross-sectional areas of 50–200 muscle fibres in different ageclasses and plotting the results as histograms with only alimited attempt at statistical analysis (Carpenè and Veggetti,1981; Stickland, 1983; Weatherley and Gill, 1985; Rowlersonet al., 1995). For example, Keissling et al. (1991) used theKolmogorov–Smirnov test statistic to provide a generalindication of differences in the frequency histograms of musclefibre cross-sectional area with growth in rainbow trout.


101 days post-hatch, January 1995


Prob

abili

ty d

ensi

ty f

unct

ion





A

B

C

D

E

F

Diploids Triploids

Diploids Triploids

Diploids Triploids

Diploids Triploids

Diploids Triploids

Diploids Triploids

0.06

0.04

0.02

0

0.06

0.04

0.02

0

0.03

0.02

0.01

0

0.03

0.02

0.01

0

0.03

0.02

0.01

0

0.03

0.02

0.01

0

0.03

0.02

0.01

0

0.03

0.02

0.01

0

0.03

0.02

0.01

0

0.03

0.02

0.01

0

0.03

0.02

0.01

0

0.03

0.02

0.01

0

0 10 20 30 40 50 0 10 20 30 40 50

0 50 100 150 200 0 50 100 150 200

0 50 100 150 2000 50 100 150 200

0 50 100 150 0 50 100 150

0 50 100 150 2000 50 100 150 200

0 50 100 150 200 0 50 100 150 200

Fig. 19. Mean value (red) andvariability band (shaded blue area) ofthe probability density functions ofmuscle fibre diameter in normal-sex-ratio Atlantic salmon 101–839 dayspost-hatch.

2012

However, when the detailed shape of the density function ofa continuous variable, such as muscle fibre diameter, is ofinterest, non-parametric smoothing techniques have distinctadvantages over methods dependent on a histogram of therealisations. The basic construction of a histogram involvesdividing the sample space into intervals and then representingeach realisation in an interval by an equal-sized box that isstacked within an interval. The result is that the function is notsmooth and the value of the realisation is lost. The kernelmethod has the advantage of preserving the value of therealisation and providing smooth density estimates that makecomparisons among different groups easier using graphicaltechniques.

To compare the distribution of fibre size between groups, wefirst estimated the mean PDF for each group using the kernelmethod. To evaluate the evidence of a structural characteristic,we graphically created a variability band using the bootstrap

technique. If the characteristic was apparent in the variabilityband, we concluded that there was evidence that thecharacteristic was genuine.

The next step was to compare the densities for diploids andtriploids within each age class. First, we used aKolmogorov–Smirnov-based bootstrap test to test the nullhypothesis of no difference. Since the Kolmogorov–Smirnovtest provides few clues to where differences might occur, wecompared the two estimated densities graphically if the testproved significant. First, we plotted the individual densitiesand then we plotted the estimated density under the nullhypothesis and placed a reference band around this estimate.The region where one or both of the group densities lie outsidethe reference band suggests where there is a major differencebetween the two densities.

Next, we compared the values of specific percentiles ofthe estimated fibre densities for each fish. The Wilcoxon








Prob

abili

ty d

ensi

ty f

unct

ion

A

B

C

D

E

Diploids Triploids

Diploids Triploids

Diploids Triploids

Diploids Triploids

Diploids Triploids

0.03

0.02

0.01

0

0.03

0.02

0.01

0

0.03

0.02

0.01

0

0.03

0.02

0.01

0

0.03

0.02

0.01

0

0.03

0.02

0.01

0

0.03

0.02

0.01

0

0.03

0.02

0.01

0

0.03

0.02

0.01

0

0.03

0.02

0.01

0

0 50 100 150 0 50 100 150

0 50 100 150 200 250 300 0 50 100 150 200 250 300

0 50 100 150 200 0 50 100 150 200

0 50 100 150 200 250 0 50 100 150 200 250

30020010003002001000

Fig. 20. Mean value (red) andvariability band (shaded blue area)of the probability density functionsof muscle fibre diameter in all-female Atlantic salmon 315–839days post-hatch.


two-sample non-parametric test was used to test whether themedian value of the specified percentile was equivalentbetween groups. Although we made multiple comparisonsfor all age classes, this stage of multiple comparisons shouldnot be scrutinised unless the Kolmogorov–Smirnov teststatistic was significant. We have included reference bandsand the results from the Wilcoxon test for the age classeswhere there was no significant difference between diploids

and triploids to demonstrate the consistency in resultsbetween the steps.

Comparing distributions between diploids and triploids byusing estimated smooth density curves alongside non-parametric tests provided us with a clearer picture of thedifferences and similarities between diploids and triploids. ForNSR fish, there were approximately one-third more musclefibres in diploid than in triploid groups at first feeding, although

101 days post-hatch 315 days post-hatch




Prob

abili

ty d

ensi

ty f

unct

ion

B

D

FE

C

A0.05

00.01

0.03

00.005

0.015

0.025

0

0.01

0.02

0.03

00.005

0.015

0.025

00.005

0.015

0.025

00.005

0.015

0.025

0 10 20 30 40 50 0 50 100 150 200

150100500200150100500

0 50 100 150 200 0 50 100 150 200

Fig. 21. Reference band (shadedarea) representing the variabilityband for the combined probabilitydensity functions of muscle fibrediameter in diploid (broken line) andtriploid (solid line) normal-sex-ratioAtlantic salmon 101–839 days post-hatch.


Prob

abili

tyde

nsity

func

tion



839 days post-hatch

B

D

E

C

A

200150100500

0.025

0.015

0.0050

0.025

0.015

0.0050

0.025

0.015

0.0050

0.025

0.015

0.0050

0

0.01

0.02

0.03

0 50 100 150

0 100 200 300

0 50 100 150 200 250 300

0 50 100 150 200 250

Fig. 22. Reference band (shadedarea) representing the variabilityband for the combined probabilitydensity functions of muscle fibrediameter in diploid (broken line) andtriploid (solid line) all-femaleAtlantic salmon 315–839 days post-hatch.

2014

the overall shapes of the probability density functions of fibrediameter were similar (P=0.32, Kolmogorov–Smirnovstatistic; Table 4). The average (not illustrated) or median fibrediameter (Table 5) provides a poor index of hypertrophicgrowth, e.g. in diploids it was relatively constant between 315days and 620 days post-hatch (Table 5) during which periodthe mean fork length increased from 14 to 25 cm (Fig. 8). Thepeak fibre diameter remains close to 20 µm in all age classesof Atlantic salmon studied because the hypertrophy of existingfibres was counterbalanced by new myotube formation. ForNSR fish, the fiftieth percentiles of fibre diameter were 12.0 %,24.5 %, 34.2 %, 20.9 % and 14.7 % greater in triploid than indiploid fish at 315 days, 470 days, 620 days, 775 days and 839days, respectively (Table 5). In contrast, Suresh and Sheehan(1998) found that there was no difference in mean fibrediameter in diploid and triploid rainbow trout over the sizerange 3–30 cm. We found evidence for a bimodal distributionof fibre diameter 620 days, 775 days and 839 days post-hatch,consistent with seasonal cycles of recruitment and expansionof the muscle fibres (Fig. 17). Carpenè and Veggetti (1981)also found evidence for seasonal cycles of recruitment inMugilidae. From an analysis of fibre size distribution in therainbow trout, Kiessling et al. (1991) inferred an increasedimportance of hypertrophy during periods of rapid growth. Inthe present study, the number of muscle fibres in Atlanticsalmon was found to increase approximately 30-fold during thefreshwater phase of the life cycle, reaching 150 000–180 000in seawater-adapted smolts. The hypertrophy of these fibres isresponsible for a significant proportion of early seawatergrowth. The left-hand peak of fibre diameters in fish sampledduring March and May 1997 probably represents fibresrecruited during the first winter in sea water. Muscle fibrenumber increased a further 3.5-fold after the first year ofseawater growth (Fig. 15A,B).

Muscle fibre densities in diploid and triploid Atlantic salmonbecame much more similar 839 days post-hatch (Tables 4–6;Figs 21, 22). The largest muscle fibres found in salmon werein the region of 200 µm diameter, which may represent theupper limit for diffusion of metabolites from the capillaries.Similar findings have been reported for rainbow trout, in whichalmost identical frequency histograms of muscle fibre diameterwere found in diploid and triploid fish greater than 30 cm infork length (Suresh and Sheehan, 1998). In the present study,similar results were obtained for AF and for NSR groups, butthe statistical evidence was less strong. Further studies with alarger number of families are required to determine whetherthere are differences in the muscle growth patterns in NSR andAF fish. Our general conclusion from this study is that thelower density of satellite cells in triploid than in diploid salmonwas associated with a proportional decrease in the rate ofmuscle fibre recruitment. Interestingly, although fibrerecruitment was compromised in triploids, there was asignificant compensatory increase in muscle fibre hypertrophy.Presumably, the lower density of satellite cell nuclei intriploids was compensated for by an increase in mean nuclearsize so that the nuclear-to-cytoplasmic volume ratio remained

relatively constant, and this may be a critical factor inhypertrophic growth. Ploidy manipulation promises to be animportant tool for dissecting the cellular and molecularmechanisms underlying hypertrophic and hyperplastic growthprocesses in fish.

The distribution of muscle fibre size and associated changesin connective tissue matrix are thought to be majordeterminants of texture and processing characteristics in fishand, as such, are important determinants of flesh quality(Hurling et al., 1996; Johnston, 1998). From a commercialstandpoint, it is interesting to note that, for these particular fish,as size increases towards the commercial slaughter body massof 2.5–4.0 kg, the distribution of muscle fibre sizes becamemore similar in diploids and in triploids, which argues againstany major differences in textural characteristics.

This study is part of a comprehensive evaluation of the useof sterile triploid Atlantic salmon supported by a FAIRcontract (AIR 3CT94 2216) from the European Commission.We are grateful for the skilled assistance of Jane Godolphinand Irvine Davidson and to Phillip McLachlan and MikeMiles and their staff at Aultbea and Almondbank for fishhusbandry.

ReferencesAliah, R. S., Yamaoka, K., Inada, Y. and Taniguichi, N. (1990).

Effects of triploidy on tissue structure of some organs in ayu.Nippon Suisan Gakkaishi 56, 569–575.

Allendorf, F. E. and Thorgaard, G. H. (1984). Tetraploidy and theevolution of salmonid fishes. In Evolutionary Genetics of Fishes(ed. B. J. Turner), pp. 1–53. New York: Plenum Press.

Benfey, T. J. (1991). The physiology of triploid salmonids in relationto aquaculture. In Proceedings of the Atlantic Canada Workshopon Methods for the Production of Non-Maturing Salmonids.Dartmouth Nova Scotia (ed. V. A. Pepper). Can. Tech. Rep. Aquat.Sci. 1789. 152pp.

Blagden, C. S., Currie, P. D., Ingham, P. W. and Hughes, S. M.(1997). Notochord induction of zebrafish slow muscle mediated bySonic hedgehog. Genes Dev. 11, 2163–2175.

Bowman, A. W. and Azzalini, A. (1997). Applied SmoothingTechniques for Data Analysis. The Kernel Approach with S-PlusIllustrations. Oxford: Oxford Science Publications, OxfordUniversity Press. 193pp.

Brown, N. C. and Stickland, N. C. (1993). Satellite cell content inmuscles of large and small mice. J. Anat. 183, 91–96.

Carpenè, E. and Veggetti, A. (1981). Increase in muscle fibres in thelateralis muscle (white portion) of Mugilidae (Pisces, Teleostei).Experientia 37, 191–193.

Carter, C. G., McCarthy, I. D., Houlihan, D. F., Johnstone, R.,Walsingham, M. V. and Mitchell, A. I. (1994). Foodconsumption, feeding behaviour and growth of triploid and diploidAltantic salmon, Salmo salar L. parr. Can. J. Zool. 72, 609–617.

Cornelison, D. D. W. and Wold, B. J. (1997). Single-cell analysisof regulatory gene expression in quiescent and activated mouseskeletal muscle satellite cells. Dev. Biol. 191, 270–283.

Davison, A. C. and Hinkley, D. V. (1997). Bootstrap Methods andtheir Applications. Cambridge: Cambridge University Press.



Devoto, S. H., Melancon, E., Eisen, J. S. and Westerfield, M.(1996). Identification of separate slow and fast muscle precursorcells in vivo, prior to somite formation. Development 122,3371–3380.

Dominov, J. A., Dunn, J. J. and Miller, J. B. (1998). Bcl-2expression identifies an early stage of myogenesis and promotesclonal expansion of muscle cells. J. Cell Biol. 142, 537–544.

Gibbons, J. D. (1971). Nonparametric Statistical Inference. London:McGraw-Hill Book Company.

Gibson, M. C. and Schultz, E. (1983). Age-related differences inabsolute numbers of skeletal muscle satellite cells. Muscle & Nerve6, 574–580.

Gonzatti-Haces, M., Seth, A., Park, M., Copeland, T., Oroszlan,S. and Vande woude, G. F. (1988). Characterisation of the trp-met oncogene p65 and the met proto-ongogene cell surface p140tyrosine kinases. Proc. Natl. Acad. Sci. USA 85, 21–25.

Greenlee, A. R., Kersten, C. A. and Cloud, J. G. (1995). Effects oftriploidy on rainbow trout myogenesis in vitro. J. Fish Biol. 46,381–388.

Grounds, M. D., Garrett, K. L., Lai, M. C., Wright, W. E. andBeilharz, M. W. (1992). Identification of skeletal muscle precursorcells in vivo by use of MyoD and myogenin probes. Cell TissueRes. 267, 99–104.

Happe, A., Quillet, E. and Chevassus, B. (1988). Early life historyof triploid rainbow trout (Salmo gairdneri Richardson).Aquaculture 71, 107–118.

Higgins, P. J. and Thorpe, J. E. (1990). Hyperplasia andhypertrophy in the growth of skeletal muscle in juvenile Atlanticsalmon, Salmo salar L. J. Fish Biol. 37, 505–519.

Hurling, R., Rodell, J. B. and Hunt, H. D. (1996). Fibre diameterand fish texture. J. Texture Stud. 27, 679–685.

Ihssen, P. E., McKay, L. R., McMillan, I. and Phillips, R. B.(1990). Ploidy manipulation and gynogenesis in fishes: cytogeneticand fisheries applications. Trans. Am. Fish. Soc. 119, 696–717.

Johnson, K. R., Wright, J. E., Jr, and May, B. (1987). Linkagerelationships reflecting ancestral tetraploidy in salmonid fish.Genetics 116, 579–591.

Johnston, I. A. (1999). Muscle development and growth in fish:potential implications for flesh quality. Aquaculture (in press).

Johnston, I. A., Cole, N. J., Abercromby, M. and Vieira, V. L. A.(1998). Embryonic temperature modulates muscle growthcharacteristics in larval and juvenile herring. J. Exp. Biol. 201,623–646.

Johnston, I. A. and McLay, H. A. (1997). Temperature and familyeffects on muscle cellularity at hatch and first feeding in Atlanticsalmon (Salmo salar L.). Can. J. Zool. 75, 64–74.

Johnston, I. A., Vieira, V. L. A. and Abercromby, M. (1995).Temperature and myogenesis in embryos of Atlantic herringClupea harengus. J. Exp. Biol. 198, 1389–1403.

Johnstone, R. and MacLachlan, P. M. (1994). Further observationson the sex inversion of Atlantic salmon, Salmo salar, using 17αmethyl testosterone. Aquacult. Fish. Mgmnt 25, 855–859.

Johnstone, R., McLay, H. A. and Walsingham, M. V. (1991).Production and performance of triploid Atlantic salmon inScotland. Can. Tech. Rep. Fish. Aquat. Sci. 1789, 15–33.

Johnstone, R. and Stet, R. J. M. (1995). The production ofgynogenetic Atlantic salmon, Salmo salar L. Theor. Appl. Genet.90, 819–826.

Johnstone, R. and Youngson, A. F. (1984). The progeny of sex-inverted female Atlantic salmon (Salmo salar L.). Aquaculture 37,179–182.

Kiessling, A., Storebakken, T., Åsgård, T and Kiessling, K.-H.(1991). Changes in the structure and function of the epaxial muscleof rainbow trout (Oncorhychnus mykiss) in relation to ration andage. I. Growth dynamics. Aquaculture 93, 335–356.

Killeen, J., McLay, H. A. and Johnston, I. A. (1999). A quantitativescoring method used to determine the effects of temperature andmaternal migratory type on development in the trout, Salmo trutta.J. Fish Biol. (in press).

Kimmel, C. B., Schilling, T. F. and Hatta, K. (1991). Patterning ofbody segments of the zebrafish embryo. Curr. Topics Dev. Biol. 25,77–110.

Koumans, J. T. M., Akster, H. A., Booms, G. H. R., Lemmens, C.J. J. and Osse, J. W. M. (1991). Numbers of myosatellite cells inwhite axial muscle of growing fish: Cyprinus carpio L. (Teleostei).Am. J. Anat. 192, 418–424.

Leary, R. F., Allendorf, F. W., Knudsen, K. L. and Thorgaard, G.H. (1985). Heterozygosity and developmental stability ingynogenetic diploid and triploid rainbow trout. Heredity 54,219–225.

Lincoln, R. F. and Scott, A. P. (1984). Sexual maturation in triploidrainbow trout, Salmo gairdneri Richardson. J. Fish Biol. 25,385–392.

Matsumoto, K. and Nakamura, T. (1996). Emerging multipotentaspects of hepatocyte growth factor. J. Biochem. 119, 591–600.

Mauro, A. (1961). Satellite cells of skeletal muscle fibers. J. Biophys.Biochem. Cytol. 9, 493–495.

Nag, A. C. and Nursall, J. R. (1972). Histogenesis of white and redmuscle fibres of trunk muscles of a fish Salmo gairdneri. Cytobios6, 227–246.

Olson, E. N. and Klein, W. H. (1994). bHLH factors in muscledevelopment: Dead lines and commitments, what to leave in andwhat to leave out. Genes Dev. 8, 1–8.

Quillet, E., Chevassus, B. and Devaux, A. (1988). Timing andduration of hatching in gynogenetic, triploid, tetraploid and hybridprogenies in rainbow trout. Genet. Sel. Evol. 20, 199–210.

Rowe, R. W. E. and Goldspink, G. (1969). Muscle fibre growth infive different muscles in both sexes of mice. I. Normal mice. J.Anat. 132, 557–579.

Rowlerson, A., Mascarello, F., Radaelli, G. and Veggetti, A.(1995). Differentiation and growth of muscle in the fish Sparusaurata (L.). II. Hyperplastic and hypertrophic growth of lateralmuscle from hatching to adult. J. Muscle Res. Cell Motil. 16,223–236.

Schultz, E. (1996). Satellite cell proliferative compartments ingrowing skeletal muscles. Devl. Biol. 175, 84–94.

Schultz, E. and McCormick, K. M. (1994). Skeletal muscle satellitecells. Rev. Physiol. Biochem. Pharmac. 123, 214–241.

Silverman, D. (1986). Density Estimation for Statistics and DataAnalysis. New York: Chapman & Hall. 175pp.

Small, S. A. and Benfey, T. J. (1987). Cell size in triploid salmon.J. Exp. Zool. 241, 339–342.

Smith, C. K., Janney, M. J. and Allen, R. E. (1994). Temporalexpression of myogenic regulatory genes during activation,proliferation and differentiation of rat skeletal muscle satellite cells.J. Cell Physiol. 159, 379–385.

Steel, R. G. D. and Torrie, J. H. (1980). Principles and Proceduresof Statistics. New York: McGraw-Hill Book Company. 633pp.

Stickland, N. C. (1983). Growth and development of muscle fibresin the rainbow trout (Salmo gairdneri). J. Anat. 137, 323–333.

Stockdale, F. E. (1992). Myogenic cell lineages. Dev. Biol. 154,284–298.

2016

Stoiber, W. and Sänger, A. M. (1996). An electron microscopicinvestigation into the possible source of new muscle fibres in teleostfish. Anat. Embryol. 194, 569–579.

Sumpter, J. P., Scott, A. P., Baynes, S. M. and Witthames, P. R.(1984). Early stages of the reproductive cycle in virgin femalerainbow trout (Salmo gairdneri Richardson). Aquaculture 43,235–242.

Suresh, A. V. and Sheehan, R. J. (1998). Muscle fibre growthdynamics in diploid and triploid rainbow trout. J. Fish Biol. 52,570–587.

Tatsumi, R., Anderson, J. E., Nevoret, C. J., Halevy, O. and Allen,R. E. (1998). HGF/SF is present in normal adult skeletal muscleand is capable of activating satellite cells. Dev. Biol. 194, 114–128.

Thorgaard, G. H. (1986). Ploidy manipulation and performance.Aquaculture 57, 57–64.

Thorgaard, G. H. and Gall, G. A. E. (1979). Adult triploids in arainbow trout family. Genetics 93, 961–973.

Van Raamsdonk, W., Van der Stelt, A., Diegenbach, P. C., Vanden Berg, W., De Beryn, H., Van Dijk, J. and Mijzen, P. (1974).Differentiation of the musculature of the teleost Brachydanio rerio.Anat. Embryol. 153, 137–155.

Veggetti, A., Mascerello, F., Scapolo, P.-A. and Rowlerson, A.(1990). Hyperplastic and hypertrophic growth of lateral muscle inDicentrarchus labrax (L.). Anat. Embryol. 182, 1–10.

Walsh, K. and Perlman, H. (1997). Cell cycle exit upon myogenicdifferentiation. Curr. Opin. Genet. Dev. 7, 597–602.

Waterman, R. E. (1969). Development of the lateral musculature inthe teleost Brachydanio rerio: A fine structural study. Am. J. Anat.125, 457–494.

Weatherley, A. H. and Gill, H. S. (1985). Dynamics of increase in

muscle fibers in fishes in relation to size and growth. Experientia41, 353–354.

Weatherley, A. H., Gill, H. S. and Rogers, S. C. (1979). Growthdynamics of muscle fibers, dry weight and condition in relation tosomatic growth in yearling rainbow trout (Salmo gairdneri). Can.J. Zool. 57, 2385–2392.

Weinberg, E. S., Allende, M. L., Kelly, C. S., Abdelhamid, A.,Murakami, T., Anderman, P., Doerre, O. G., Gunwald, D. J.and Riggleman, B. (1996). Developmental regulation of zebrafishMyoD in normal sex ratio, no tail and spadetail embryos.Development 122, 271–280.

Wilson, S. J., Mcewan, J. C., Sheard, P. W. and Harris, A. J.(1992). Early stages of myogenesis in a large mammal: formationof successive generations of myotubes in sheep tibialis cranialismuscle. J. Muscle Res. Cell Motil. 13, 534–550.

Yablonka-Reuveni, Z. and Rivera, A. J. (1994). Temporalexpression of regulatory and structural proteins during myogenesisof satellite cells on isolated rat fibers. Dev. Biol. 164, 588–603.

Youngson, A. F., Hansen, L. P. and Windsor, M. L. (1998).Interactions between salmon culture and wild stocks of Atlanticsalmon: The scientific and management issues. In Report by theConvenors of a Symposium by the International Council forExploration of the Sea (ICES) and the North Atlantic SalmonConservation Organisation (NASCO), Bath, England, UK, April1997. Norway: Norwegian Institute for Nature Research.

Yun, K. and Wold, B. (1996). Skeletal muscle determination anddifferentiation: Story of a core regulatory network and its context.Curr. Opin. Cell Biol. 8, 877–889.

Zar, J. H. (1996). Biostatistical Analysis. London: Prentice-HallInternational. 662 pp.


Documents

Salmon muscle growththe myosepta was slightly more advanced in triploid than in diploid ﬁsh, although the differences were smaller than among individual families. The c-met receptor