Chromosome polymorphism in wild Atlantic salmon, Salmo salar L., from Asturias, northern Spain

Aquaculture and Fisheries Management 1992, 23, 95-10]

Chromosome polymorphism in wild Atlantic salmon,Salmo salar L., from Asturias, northern Spain

E. GARCIA-VAZQUEZ, A. M. PENDAS & P. MORAN Departamento de BiologiaFuncional (Area Genetica), Universidad de Oviedo, Oviedo, Spain

Abstract. Nine samples of Atlantic salnion, Salmo salar L., from four different rivers inAsturias (northern Spain) and three different life stages (fry, parr, adult) were karyotyped. Thedistributions of 2n chromosome numbers in the samples were not statistically different,suggesting the Asturian Salmo salar populations have the same chromosomal polymorphismmake-up. They were, however, significantly different from populations in Scotland andNorway used to stock many Asturian rivers.

Introduction

Atlantic salmon, Salmo salar L., is a species of considerable interest to anglers and scientists.Its geographic distribution has as its southern limits the rivers of northern Spain, wherepopulations are as yet poorly studied. Only in recent years has relevant research beenpublished (Garcia-Vazquez, Linde. Blanco, Sanchez, Vazquez & Rubio 1988; Garcia deLeaniz & Martinez 1988; Martin Ventura 1988; Blanco, Sanchez, Vazquez, Garcia &Rubio 1990). In populations from Asturias (Fig. 1). the life cycle is commonly 1 year in theriver and 2 years in the sea (Martin Ventura 1988). This life cycle is shorter than that of fish innorthern latitudes, probably due to a more productive environment and warmer freshwatertemperatures. These data point to the possibility that Asturian populations could perhaps bepeculiar genetically, both at the biochemical and karyotypical levels due to adaptation totheir unique environmental circumstances. Genetic characterization is an important step inthe conservation of these populations.

Chromosome polymorphism is employed in salmonids to characterize populations (Gold& Gall 1975; Thorgaard 1983; Garcia-Vazquez et al. 1988; Moran, Pendas, Garcia-Vazquez & Linde 1989) by intrapopulationa! variation in the number of chromosomeskeeping the number of chromosomic arms (NF) unchanged. The patterns of chromosomepolymorphism are different in the different populations of Salmo salar studied (Roberts1968; Grammeltvedt 1975; Barshene 1978; Hartley & Horne 1984; Garcia-Vazquez et al.1988); this is the method used in this study to characterize Asturian populations.

Materials and methods

In this st\jdy, karotypes of salmon in three life stages (fry, parr and adult) were analysedduring three different years (1987,1988 and 1989) from four rivers in Asturias: Cares, Sella,

Correspondence: Dr Eva Garcia-Vizquez, Departamento de Biologia Funcional (Area Genetica), Universi-dad de Oviedo, 33071 Oviedo, Spain.

95

96 E. Garcia-Vdzquez et al.

FRANCE

Figure 1. Geographic map of the most important rivers in Asturias (northern Spain).

Narcea and Esva (Fig. 1). Samples are distributed as follows: three fry, two parr and fouradult samples.

Fry. Fry were obtained in 1987 (PrN-87) by spawning of 21 adults captured in the Narcea(8 females and 13 males); 6 females and 12 males in 1988 (FrN-88); 8 females and 12 males in1989 (FrN-89). Aloctonous fry employed to repopulate the rivers of Asturias in the 3 yearswere also analysed. These fry were imported from a Scottish hatchery.

Karyotypes were obtained in one-month-old fry. The method used for obtainingchromosome preparations follows that of Chourrout (1984).

Parr. The samples were obtained by electrofishing twice in the same area of the Esva: in1987 (PaE-87) and in 1989 (PaE-89). Parr were 0+ and 1+ taken prior to smolt run.

Karyotypes of parr were obtained in 1987 following Chourrout & Happe (1986), and in1989 as described in Garcia-Vdzquez, Moran & Pendas (1991).

Adults. The samples were collected in 1989 in those four Asturian rivers mentionedabove: Cares (AdC-89), Sella (AdS-89), Narcea (AdN-89) and Esva (AdE-89).

Karyotypes of adults were obtained by lymphocyte culture following Hartley & Horne(1982, 1985).

At least five metaphases per individual were counted.

Results

Table 1 gives the number of individuals of each chromosomic number per sample. Theintraindividual chromosome polymorphism observed is smaller than 0-5% and it is not taken

Chromosomes of Atlantic salmon from Asturias 97

into account when analysing data. The mode of chromosomic numbers is 2n = 58 in the ninesamples analysed (Table 1) with a constant NF = 74, a typical value in European salmon(Hartley 1987). Chromosomic numbers higher than 59 and lower than 56 have not beenfound.

There are no significant differences among all the Asturian samples (x^ = 23-76,16 df, notsignificant). A joint chromosome distribution of the whole of the Asturian samples, calledAsturian distribution (Fig. 2). is meant to show that this distribution has a clear mode in 2n =58, a small proportion of individuals with 2n = 59 (5%) and a similar proportion of both 2n =56 and 2n = 57, close to 17%.

The populations of the four rivers analysed in the adult stage are very similar (see Table 2:X̂ = 0-128, 3 df, NS), without 2n = 59 individuals.

In the Esva. the chromosome distributions of the three samples analysed (parr of 1987 and1989, and adults of 1989) are not different at a statistical level (x^ = 2-62, 4 df, NS). Bothstages of the salmon's life have the same distribution, without significant changes in parr from1987 to 1989, and from parr to adult. As the sea stage in Asturian salmon often lasts 2 years(Martin Ventura 1988), about 60% of adults returning in 1989 were parr in 1987.

Table I. Number (and percentage) of individuals of each chromosomic number, in eachanalysed sample, x^ of contingency test of difference between samples (in actual numbers),adding column 2n = 59 and 2n = 58 because of small expected numbers

Sample

FrN-87FrN-88FrN-89PaE-87PaE-89AdC-89AdS-89AdN-89AdE-89

X = 23-76,

56

1 (4-8%)6(13-3%)

32(24-6%)1 (5 0%)7(15-2%)2(11-1%)4(16-0%)3 (13-0%)3(11-1%)

16df, NS

Chromosomic

57

4(191%)15 (33-3%)20(15-4%)

3 (15-0%)9 (19-6%)2(111%)2 (8-0%)3 (13-0%)3(!M%)

number (2ti)

58

14 (66-7%)18(40-0%)69(53-1%.)15(75-0%)29(63-0%)14(77-8%)19(76-0%)17 (73-9%)21 (77-8%)

59

2(9-5)%6(13-3%)9(6-9%)I (5-0%)1 (2-2%)0000

Table 2. Tests of x' of contingency measuring differences between samples. Adults: AdC-89,AdS-89, AdN-89 and AdE-89. Ast: general chromosome distribution of salmon fromAsturias. Esv: chromosome distribution of salmon from the River Esva. Nar: chromosomedistribution of salmon from the River Narcea, except fry of 1988

Samples compared

AdultsPrE-87, PrE-89, AdE-89FrN-87, FrN-88. FrN-89, AdN-89FrN-87, FrN-88, AdN-89 (Nar)Nar. FrN-88Ast. FrN-88Ast, Esv. Nar

X̂ value

0-1282-617

19-4228-151

11-21612-7856-755

df

3496336

Statisticalsignificance

NSNS

P < 0-05NS

P < 0-05P < 0-01

NS

98 E. Garcia-Vdzquez et al.

ASTESVNAR

7 0

6 0

5 0

-o

1 40

3 0

20

10

56 57 58Chromosomic number

59

Figure 2. Chromosome distributions (in percentage of individuals) as follows: AST (general chromosomedistributions of salmon from Asturias). ESV (chromosome distribution of salmon from the River Esva) and NAR(chromosome distribution of salmon from Narcea except FrN-88).

In the Narcea, chromosome distribution remained constant from year to year in fry, beingthe same as that of adults (Table 1), except in 1988, when the proportion of less commonchromosomic numbers (57 and 59) was higher. This was the only chromosome distributionthat differs significantly from the other Asturian distributions (x^ = 12-785, 3 df, P < 0-01,Table 2). FrN-88 is not taken into account for calculating the characteristic distribution ofNarcea (Fig. 2), which is more variable (the 2N = 58 is lower) than the general Asturiandistribution.

Table 3 shows chromosome distributions of aloctonous stocks employed to repopulateAsturian rivers in 1987,1988 and 1989. We can see the high proportion of 2n = 59 individualsand the low proportion of 2n = 56 individuals when comparing with Asturian samples. JointScottish chromosome distribution (SCOT) was significantly different from the Asturian one(X̂ = 42-51, 3 d f , P < 0-001).

Discussion

The very low intraindividual polymorphism found, less than 0-5% of cells analysed, is incontrast to the observations of other authors. Hartley & Hor^ie (1984) given 12-38% for this

Chromosomes of Atlantic salmon from Asturias 99

Table 3. Number (and percentage) of individuals of each chromosomic number, in thealoctonous stocks employed in Asturian rivers repoputalion along 1987, 1988 and 1989. x^ ofcontingency test (in actual numbers) of difference between Scottish and Asturian chromosomedistributions. Scot: general chromosome distribution of salmon imported from Scotland

Year

198719881989

ScotAst

Chromosomic number

56

03 (6-7%)0

3 (2-5%)59(16-6%)

57

4(14'8%)6(13-3%)4 (8-0%)

14(11-5%)61(17-2%)

58

19 (7047..)20 (44-4%)39 (78-0%)

78 (63-9%)216 (60-8%)

59

4(14-8%)16(35-5%)7(M'0%)

27(22-1%)19 (5-3%)

X̂ = 42-51, 3d f ,P< 0 001

value. Bolla (1987) found Robertsonian intra-individual polymorphism in 1% of the cells.Our data agree with the latter.

The work shows that Asturian populations have characteristic patterns of chromosomepolymorphism. Moreover, these patternsare the same over the years in spite of the fact that,obviously, each year there were different spawners. Both Esva and Narcea populations didnot change either in the years in which they were studied, or in the different life stages, exceptfor FrN-88.

We observed a variation in this distribution in fry of 1988, whose variability was higherthan in the other eight samples. This could be simply due to the small number of parents forspawning, which in 1988 was 6 females and 12 males. Ueda & Ojima (1984) studied theprogeny of a female of Oncorhynchus mykiss Walbaum 2n = 61 and found that most of thedescendants presented the maternal chromosome number. Therefore, the effect of only onefemale becomes more important as fewer females are used. One exceptional female 2n = 59should move significantly the mean distribution of offspring to higher chromosome numbers,when there are only five other females contributing to the spawning.

Every sample has a high proportion of individuals with 2n = 56, higher than the resultsreported for other populations in Europe, including Scottish stocks used for repopulatingAsturian rivers (Tables 3 and 4). Asturian samples have also higher interindividualvariability.

Chromosomes are the visible organization of the genome; in different isolated popula-tions this organization would follow differently. In Salmo salar, different chromosomeorganization (basically different DNA sequences) could be reflected in chromosomepolymorphisms. The standard European Salmo salar karyotype consists of 16 metacentricand submetacentric chromosomes and 2 large, 22 medium and 18 small acrocentrics. Thesmallest acrocentrics are involved in Robertsonian translocations (Roberts 1968, 1970;Hartley ^ Horne 1984); fusions would be between non-homologous acrocentrics to producesmall metacentrics (May, Wright & Johnson 1982; Hartley & Home 1984; Hartley 1987 forreview). Different isolated populations have very different chromosome polymorphismpatterns, with very diverse variation in chromosome numbers (Table 4) (Roberts 1970;Barshene 1978; Hartley & Horne 1984; Bolla 1987; Hartley 1988; Garcia-Vazquez et al.1988). Homologous identification for all the chromosome pairs seems to be the first step toidentify those chromosomes involved in the polymorphism. However, high resolution

100 E. Garda-Vdzquez et al.

Table 4. A percentage of individuals of each chromosomic number, in samples of Salmo salar from differentgeographic origin. All the samples showed intraindividual polymorphism. Machiasand Miramichi percentages werecalculated from cell results. NI: sample's number of individuals

Sample

Farm ScottishWild ScottishFarm ScottishSun. (Norway)Machias (USA)Mirami (USA)

Authors

Hartley &Horne (1984)Hartley (1988)Hartley (1988)Grammeltvedt (1975)Roberts (1970)Roberts (1970)

<56

--

-87-343-5

Chromosomic number

56

3-3---6-7

251

57

13-32 4

H I-3-4

193

58

83-397-688-9

1001-98-9

>58

----

3-1

Nl

3042276

3390

analysis of DNA banding for the chromosomes in Salmonids still has not been developed(Phillips, Pleyte, Ert & Hartley 1989). The polymorphisms may arise because differentchromosome pairs are involved in Robertsonian translocations in different populations. Cand Q banding studies (Phillips & Hartley 1988) provide no evidence that Robertsoniantranslocations are restricted to a few chromosomes. If so, the result of mixing populationswith different chromosomes resulting from different Robertsonian rearrangements would bean increased incidence of multivalents, or of chromosomal aberrations in meiotic divisionsand gametogenesis in hybrid descendants.

These results show a clear difference in the frequency of different 2n numbers ofchromosomes between native and non-native stocked individuals in Spain. This is likely to beimportant. The introduction of stocks with different chromosome distributions may havenegative fitness implications when non-native salmon cross with native fish. If so, it will bebest to use autoctonous spawners when stocking to maintain the existing chromosomestructure rather than to change it by introducing aloctonous stocks.

Acknowledgments

This work was supported by the FICYT and the CAICYT15/84. We wish to thank Dr Hartleyfor her suggestions and D. Armando Gines Gonzalez for his collaboration. Finally thanks tothe Agencia del Media Ambiente del Principado de Astuxias for allowing us to take samplesin Asturian rivers.

References

Barshene J.V. (1978) Ontogenetic variability of the chromosome number in the Atlantic salmon (Salmo salar L.)Genetika 14, 2029-2036.

Blanco G., Sanchez J.A., V ^ u e z E., Garcfa E. & Rubio J. (1990) Superior developmental stability ofheterozygotes at enzyme loci in Salmo salar L. Aquaculture 84, 199-209.

Bolla S. (1987) Cytogenetic studies in Atlantic salmon and rainbow trout embryos. Hereditas 106, 11-17.Chourrout D. (1984) Pressure-induced retention of second polar body and suppression of first cleavage in rainbow

trout; production of all-triploids and heterozygous and homozygous diploid gynogenetics. Aquaculture 36,lU-126.

Chourrout D. & Happe A. (1986) Improved methods of direct chromosome preparation in rainbow trout, Salmogairdneri. Aquaculture 52, 255-261.

Chromosomes of Atlantic salmon from Asturias 101

Garcia de Le4nlz C. & Martinez J.J. (1988) The Atlantic saimoti in the rivers of Spain, with particular reference toCantabria. In: Atlantic Salmon: Planning for the Future {ed. by D. Mills & D. Piggins). pp. 179-209. CroomHelm, London.

Garcia-Vazquez E., Linde A.R., Blanco G., Sfinehez J.A., V5zquez E. & Rubio J. (1988) Chromosomepolymorphism in farm fry stocks of Atlantic salmon from Asturias. Journal of Fish Biology 33, 581-587.

Gareia-VSzquez E., Morin P. & PendSs A.M. (1991) Chromosome polymorphism patterns indicate failure of aScottish stock of Salmo salar transplanted into a Spanish river. Canadian Journal of Fisheries and AquaticSciences 4S, 170-172.

Gold J.R. & Gall G.A.E. (1975) Chromosome polymorphism in the California High Sierra golden trout (Salmoaguabonita). Canadian Journal of Genetics and Cytology 17, 41-53.

Grammeltvedt A.F. (1975) Chromosome of salmon (Satmo salar) by leukocyte culture. Aquaculture 5^205-209.Hartley S.E. (1987) The chromosomes of salmonid fishes. Biological Review 62, 197-214.Hartley S.E. (1988) Cytogenetic studies of Atlantic salmon, Sa/mo5a/ar L., in Scotland./oHma/o/Fts/iflio/ogy 33,

735-740.Hartley S.E. & Horne M.T. (1982) Chromosome polymorphism in the rainbow trout {Satmo gairdneri Richardson).

Chromosoma 87, 461—468.Hartley S.E. & Horne M.T. (1984) Chromosome polymorphism and constitutive heterochromatin in the Atlantic

salmon, Salmo salar. Chromosoma 89, 377-380.Hartley S.E. & Horne M.T. (1985) Cytogenelic techniques in fish genetics. Journal of Fish Biology 26, 575-582.Martin Ventura J.A. (1988) The Atlantic salmon in Asturias, Spain. In: Atlantic Salmon: Planning for the Future

(ed. by D. Mills & D. Piggins), pp. 210-227, Croom Helm, London.May B., Wright J. E. Jr & Johnson K.R. (1982) Joint segregation of biochemical loci in salmonidae. III. Linkage

association in salmonidae including data from rainbow trout (Salmo gairdneri). Biochemical Genetics 20,29-40.

Morin P.. Pendas A.M., Garcia-Vazquez E. & Linde A.R. (1989) Chromosomal and morphological analysis oftwo populations of Salmo trutta sbp fario employed in repopulation. Journal of Fish Biology 35, 839-843.

Phillips R.B. & Hartley S.E. (1988) Fluorescent banding patterns of the chromosomes of the genus Satmo. Genome30, 193-197.

Phillips R.B.. Pleyte K.A., Van Ert L.M. & Hartley S.E. (1989) Evolution of nucleolar organizer regions andribosomal RNA genes in Salvelinus. Physiology and Ecology of Japan Spec. Vol. I, 429-447.

Roberts F.L. (1968) Chromosomal polymorphisms In North American landlocked Salmo salar. Canadian Journal ofGenetics and Cytology 10, 865-875.

Roberts F.L. (1970) Atlantic salmon (Satmo satar) chromosomes and speeiation. Transactions of the AmericanFisheries Society 99, 105-111.

TTiorgaard G.H. (1983) Chromosomal differences among rainbow trout populations. Copeia3, 650-662.UedaT. & Ojima Y. (1984) Cytogenetical characteristics of the progeny from the heteroploidy in the rainbow trout.

Proceedings of the Japanese Academy Serie B 60, 183-186.