Genome size estimation in two populations of the Northern Chilean scallop, Argopecten purpuratus, using fluorescence image analysis.

ABSTRACT Genome size (the C-value) is known to vary considerably among organisms, but is relatively constant among individuals of the same species. To determine the genome size in the Northern Chilean scallop Argopecten purpuratus, two populations were analyzed by measuring the fluorescence signal in hemocyte nuclei stained with DAPI. The fluorescence intensity was measured during a fluorescence-fading period by image analysis (fluorescence fading method). The area under the curve during a fading period was associated with the genome size, thus we determined the C-value in the Arica population as 1.057 [+ or ] 0.057 pg of DNA and 1.139 [+ or -] 0.066 pg of DNA for the Tongoy population. The interindividual variation was smaller that 6% for both populations. Variance analysis was performed to detect the effects of organisms and population in the genome size of A. purpuratus. Significant differences were detected between Arica and Tongoy populations (P = 0.0011). The genome size of Tongoy population was larger statistically compared with the Arica population. Several hypotheses are discussed about genome size variation.

KEY WORDS: genome size, Argopecten purpuratus, image analysis, fluorescence fading

INTRODUCTION

The Northern scallop Argopecten purpuratus (Lamarck, 1819) is a functional hermaphrodite inhabiting sedimentary substrates in sheltered bays along the Pacific Ocean from Paita, Peru (5[degrees] South) to Tongoy, Chile (30[degrees] South) (yon Brand-Skopnik & Ibarra-Humphries 2002). Since the 1980s, A. purpuratus has been cultured successfully and currently represents an important aquaculture product in Chile (Gajardo & Nunez 1992). Because of its economic importance in Northern Chile, several studies have been carried out to enhance the production of this scallop species (Navarro et al. 2000, Martinez & Perez 2003). Furthermore, the management reached in hatchery and larvae production has permitted to apply methodologies like ploidy manipulation to produce triploid organisms (Canello et al. 1992, Winkler et al. 1993) analyzed, using the chromosome count described by yon Brand et al. (1990). However, the knowledge about the genetic characteristics in populations of A. purpuratus along the Pacific coast is sparse (von Brand & Kijima 1990, yon Brand-Skopnik & Ibarra-Humphries 2002).

Genome size analyses are important to define genetic characteristics of a species and to estimate the ability of that species to undergo evolutionary changes. This type of analysis may be especially useful in biodiversity and conservation studies, as well as determination of ploidy levels in genetically manipulated species (Hinegardner 1974, Thorgaard et al. 1982, Komaru et al. 1988, Rodriguez-Juiz et al. 1995).

Diverse relationships on the haploid genome size (C-value) have been studied to determine the evolutionary role of the genome size (Gregory 2001 a). However, the genome size implication and its evolutionary role is a very complex topic, because of the variability of the genome sizes among eukaryotic species (Gregory 2002). The eukaryotic genome size varies more than 200,000-fold (Gregory 2001b). Nevertheless, whereas genome size varies significantly among species, the gene number varies considerably less.

Several methods have been used to quantify genome size. Densitometric techniques and flow cytometry have been used mainly because they allow to obtain estimations of genome size in many samples in a short time. However, the possible disadvantage of these techniques involves the Feulgen stain protocol in densitometry and the requirement of nuclei in suspension in flow cytometry. Recently, image analysis methods have been introduced in cytogenetic research (Uozu et al. 1997). Furthermore, we reported a new method for genome size estimation based on the fluorescent decay lifetime when the fluorochrome is exposed to excitation light. According to our results (Gallardo-Escarate et al. 2004), the DNA-fluorescent fading shows a relationship with the nuclear DNA content. To establish a fading unit we measured the fluorescence intensity during a blanching period using image analysis. The area under the curve (IF) was related with the genome size of several species. Our results showed that the IF in spermatozoa and red blood cell nuclei of Oreochromis mossambicus, spermatozoa of Haliotis rufescens, and red blood cell nuclei of Oncorhynchus mykiss showed a linear relationship with genome size reported for these species (r = 0.99), therefore we proposed a linear equation:

y = 3358.8x + 801.82

to estimate other genome sizes. This work demonstrated the feasibility to estimate genome sizes by quantification of fluorescent fading.

In this study, we applied the method mentioned earlier for genome size estimation in two populations of the northern Chilean scallop Argopecten purpuratus.

MATERIALS AND METHODS

Biologic Material

Thirty organisms of Argopecten purpuratus were used to estimate the genome size in two scallop populations. Fifteen scallops were collected from a wild population in Arica (18[degrees]28'S) and 15 were collected from the cultured population in Tongoy Bay (30[degrees]15'S), both areas are located in the north of Chile. Shell length (SL), width (SW), height (SH) of each scallop were measured and SW/SL, SH/SL ratios were obtained.

Cell Preparation and DNA-fluorescence Stain

Genome size was estimated using hemocyte nuclei of adult scallops (8-9 cm SH). The hemolymph was extracted from the adductor muscle of living animals using an insulin syringe. Hemolymph was smeared and fixed on clean slides with fresh Carnoy (Methanol: acetic acid, 3:1) at 4[degrees]C and air-dried. Fixed cells on slides were washed in 3 changes with Phosphate Buffer Saline (13 mM NaCl, 0.2 mM KCl, 0.8 mM [Na.sub.2]HP[O.sub.4], 0.2 mM K[H.sub.2]P[O.sub.4], pH 7.4) (PBS1X) for 5 min each time, and incubated with DAPI solution in the dark for 25 rain at room temperature (25[degrees]C) using a coplin jar. The DAPI solution stain was prepared with 4,6-diamidino-2-pbenylindole (DAPI) (Sigma-Aldrich, USA) in PBS1X at 0.5 [micro]g/mL. After that slides were observed only with PBS1X, neither antifading solution was used.

Image Capture and Fluorescence Fading Analysis

The fluorescence was measured using a motorized epifluorescent microscope (Leica model DMRXA2) equipped with a RGB color digital camera of 36-bits and 3.3 mega pixels (Leica model DC300). The fluorescence conditions were: excitation filter UV 340-380 nm, dichromatic mirror DM 400 nm, suppression filter LP 425, absorption band AB 435-485 nm, and objective lens PL Fluotar 63 x / 0.7. The conditions for image capture were: CCD exposure time of 509 milliseconds, gamma equal to 1 (linear response) and color depth of 8 bit/channel. The image capture was generated by programming the digital capture process using QWIN software (Copyright by Imaging Systems Ltd, Cambridge, UK, 1997). This program captured a digital image of 348 x 258 pixels each 1.6 sec during fluorescent fading period on the same observational field. Additionally, the program stored the captured images in a folder for its analysis.

The image analysis was performed by an algorithm specifically built by us using MATLAB software (Copyright 1984-2000, The MathWorks, Inc.). The algorithm allowed to estimate the area under the curve by one integration function of the pixel intensities (fluorescence intensity) in 256 level of brightness, thus for each nucleus DNA-fluorescent the mean intensity in a 256 scale was calculated. The pixel mean intensity (PMI) was estimated in a window of 13 x 13 pixels centered in the nucleus. Each nucleus within the image was simultaneously measured during fluorescence fading period until the fading time was completed. If all PMI were plotted during a fluorescent fading period, then the area under the curve represents the IF like:

(1) IF = [n.summation over (i=1)] [(PMI).sub.i] * [DELTA]t

where the subscript i represents the number of images in a fading period and [DELTA]t the time among each image captured. Thus IF represents the integral fading by the sum of the mean intensity (fluorescence) during the fading period. The C-value in Argopecten purpuratus was obtained by means of the linear model built with standard C-values. To estimate genome size in A. purpuratus, we used the linear equation 33 y = 3358.8x + 801.82. Where y = IF value obtained from hemocytes and x = unknown nuclear DNA content (pg). The linear equation and standard error were calculated by Fisher analysis.

Statistical Analysis

In reference to SL and SW/SL and SH/SL, ratios were compared between populations with t-tests each. To obtain the intraspecific and interpopulational genome size variation, several analyses were carried out to estimate the percentage of variation at each sampling level (individuals and populations), and its contribution to the total genome size variation was detected. The coefficient of variation was estimated from distribution of DNA values of individuals, and analysis of variance were performed using 1-way and 2-way nested. The ANOVA assumptions, normal distribution, and homogeneity of variances were tested using Kolmogorov-Smirnov test and Bartlett test respectively. The statistical analysis was carried out using Statistica 6.1 software (Copyright 1984-2002, StatSoft, Inc.).

RESULTS

Morphologic Measurements

Scallops mean SL were 7.69 ([+ or -] 1.290) and 6.97 ([+ or -] 0.481) cm for Tongoy and Arica populations respectively, whereas SH/SL ratios were 1.063 ([+ or -] 0.027) and 1.064 ([+ or -] 0.021), and SW/SL ratios were 0.339 ([+ or -] 0.019) and 0.443 ([+ or -] 0.023), respectively. There were no significant differences either between mean SL [(t.sub.(17)] : 2.03, P = 0.058) or mean SW/SL ratios ([t.sub.28] = -0.12, P = 0.90) of both populations. However, Tongoy population mean SH/SL ratio was statistically smaller than Arica population mean SH/SL ratio ([t.sub.28]) = -13.50, P < 0.0001).

Fluorescence Fading and Genome Size

The brightness in DAPI-stained hemocyte nuclei was analyzed to determine the fluorescence intensity. The fluorescence in A. purpuratus hemocytes was recorded in color digital images of 8 bits. This kind of image allows converting the RGB (true color) image to pseudocolor images. The pseudocolor image allows visualizing and quantifying the fluorescence intensity by means of 256 levels of brightness (Fig. 1). Furthermore, the fluorescence intensity was measured in a window of 13 x 13 pixels centered in the nucleus. This window was used to estimate the PMI in hemocyte nuclei during a fluorescent decay lifetime (Fig. 2). The highest PMI was found at the beginning of the exposition to ultraviolet light, whereas the lowest PMI was found at the end of the exposition.

[FIGURES 1-2 OMITTED]

In this study we used the fluorescent fading to estimate the genome size (C-value) in two populations of A. purpuratus. Profile of fading function (mean [+ or -] SD) was generated by fluorescence intensity quantification. The area under the curve represents the IF (Eq. 1). The fluorescence fading was measured each 1.6 sec until total fading in hemocytes of A. purpuratus (Fig. 3).

[FIGURE 3 OMITTED]

Genome Size Variation

The Arica population showed a minimum C-value of 0.932 pg and a maximum of 1.179 pg of DNA. The total number of nuclei measured was 5,628. The individual variation (standard deviation) was smaller than 0.32 pg of DNA in all scallops sampled. The Tongoy population instead, showed a minimum genome size of 1.044 pg of DNA and a maximum value of 1.285 pg of DNA in 7,538 hemocyte nuclei measured. The individual variation was between the 0.105 pg and 0.397 pg of DNA (Table 1). The interindividual (n = 15) variation was smaller than 6% for both populations (coefficient of variation). For both populations, the genome size variation showed to be steady in variance and mean near to 1,000 measured nuclei (Fig. 4, Fig. 5). In this context, it is possible to obtain a representative C-value in a sample size of 150 nuclei measured per organism.

[FIGURES 4-5 OMITTED]

The using of image analysis allows to analyze a large number of hemocyte nuclei (n = 13,166) and therefore determine the C-value for both populations. The Arica population showed a genome size of 1.057 [+o r -] 0.057 pg of DNA. The confidence interval (95%) was 1.026 [less than or equal to] C-value [less than or equal to] 1.089 pg of DNA. The Tongoy population showed a C-value of 1.139 [+ or -] 0.066 pg of DNA and the confidence interval (95%) was 1.102 [less than or equal to] C-value [less than or equal to] 1.175 pg of DNA.

Analysis of variance was performed to detect the effects of organisms and population in the genome size of A. purpuratus. Significant differences were detected between Arica and Tongoy populations (P = 0.0011). The genome size of Tongoy population was larger statistically compared with Arica population, 1.139 [+ or -] 0.066 pg and 1.057 [+ or -] 0.057 pg respectively. The contribution of the different sampling levels (individuals, population, and error) to the total variance in genome size was analyzed by 2-way ANOVA nested (Table 2). The subdivision of percentages of variance attributed to each level shows that the variance error (measurements within an individual) represent <1% of the total. The most of the variation in genome size occurs between populations. The analysis of the interindividual variation showed that significant differences exist and its variation was near to 6% of DNA content for both populations.

DISCUSSION

Although the accurate measurement of genome sizes is not simple, it has enough practical and theoretical importance to justify the effort. Recent advances in computing and image analysis technology has allowed to obtain a fast and reliable estimation of genome size (Hardie et al. 2002). Image analysis densitometry has been accepted as an accurate means of quantifying DNA for diverse applications (Bertino et al. 1994, Gregory 2003) and its broader utility in the measurement of both plant and animal genomes (Hardie et al. 2002). However, the disadvantage of this methodology resides basically in two aspects: the staining protocol and its quantifying method to obtain an accurate measurement of genome size.

In fluorescence observation by conventional fluorescence microscopy, the retardation of fluorescence fading, the high initial intensity of images, and low background noise are important factors for obtaining accurate images (Ono et al. 2001). DAPI stain is the most widely used fluorochrome to stain nuclear DNA in fluorescence microscopy. However, the fluorescence of DAPI is rapidly lost when exposed to excitation by ultra violet light especially under optimal condition for observation, under illumination at the wavelength of maximal absorbance of the fluorochrome, and with objective lenses of a high numerical aperture (Johnson & Nogueira-Araujo 1981). The photochemical process underlying the fluorescence decay has not yet been fully explained, although some theories suggest the involvement of oxygen, triplet states, and protein denaturalizations (Hirschfeld 1979, Johnson et al. 1982). However, to prevent retardation of fading in microscopy several methods can be used to remove oxygen from the mounting medium. Furthermore, it is feasible to increase the viscosity of the medium and avoid the oxygen diffusion (Johnson et al. 1982, Ono et al. 2001). The point is to retard the diffusion of oxygen and consequently reduce the amount of oxidized fluorochromes. Our proposed method only used phosphate buffer saline (PBS) as the mounting medium, therefore the photochemical process is not modified. According to our results, we found evidence that the fading is a function of nuclear DNA content in hemocytes, because in all analyzed nuclei these showed a uniform variation to ultra violet light exposition after, and the IF value measured have relationship with the lineal model above. The present method by analysis of fluorescence in digital images permits to obtain a fast and accurate method for DNA content quantification and together represent an alternative method to flow cytometry.

Genome size (the C-value) is known to vary considerably among organisms, but it is relatively constant among individuals of the same species. Its significant degree of variation among organisms is more than 200,000-fold among eukaryotes (Gregory 2001a), but it is not related to biologic complexity. Several hypotheses have been proposed to explain this genome size variation.

The "selfish DNA" hypothesis suggests that the accumulation of DNA over evolutionary time may result from the cell replication process. The increase in genome size thus ceases when it begins to compromise cell viability (Orgel & Crick 1980). According to this theory, larger cells can simply tolerate more DNA (Gregory 2000). An alternative hypothesis postulates that genome size may influence the fitness of the organisms by a nucleoskeletal effect that is based on the amount of nuclear DNA and is independent of the nucleotide sequence. In this sense, genome size is considered to be a selective trait, because larger cells will need larger nuclei to preserve a favorable metabolic balance (Cavalier-Smith & Beaton 1999).

According to our results, significant differences were found in the genome size of A. purpuratus in the two populations studied (P = 0.0011). The Tongoy population has a larger genome size than Arica population, with mean genome size of 1.139 [+ or -] 0.066 pg and 1.057 [+ or -] 0.057 pg respectively. According to Dolezel et al. (2003), 1 picogram (pg) equals 0.978 x [10.sup.9] bp, and then the difference between both populations was 8.019 x [10.sup.7] bp. However, the interindividual genome size variation was similar to 6% of DNA content for both populations. In this context, Rodriguez-Juiz et al. (1995) reported a significant heterogeneity in mean genome size between individuals within species, as well as among different species of bivalve molluscs. A high percentage of the total genome size variation was allocated at the species level (92%) and 7% at the level of individuals within species. These data and our results (6%) suggest that a significant fraction of the mollusc's genome is free to vary among individuals of the same species without phenotypic or biologic consequences. Furthermore, the genome size reported in our study for A. purpuratus populations was close to Argopecten irradians C-value (1.20 pg) and within the interval reported for other Pectinid species (Table 3).

The farming of the Chilean scallop was greatly promoted by the fact that natural beds were almost depleted. Nowadays, a fast growing industry is still far from realizing its potential because of the limited, or unreliable, production of quality seeds (Gajardo & Nunez 1992). This work offers an alternative step to compare probable differences among the few wild population remaining and hatchery stocks. The differences found in genome size of A. purpuratus may be due to wild and cultivated origin of the populations studied. The organisms from Tongoy population have been exhaustively farmed in the last 3 decades. Further, the management of the larvae production and the seed selection method probably has originated a selection process toward a specific genotype (domestication). In this context, in a recent study of Drosophila population, Vieira et al. (2002) found that changes in genome size produced by the activation of transposable elements (TEs) were associated with the invasion of new habitats. The main point that emerged from these findings is that the TE insertion site number on chromosome arms (the euchromatic copies) only contributes to ~5% to 9% of the observed genome size variation. This suggests that genome size variation is caused by other factors as well as by TEs in euchromatin. However, limited evidence has been provided to support the existence of an approximately linear relationship between total TEs and genome size (Kidwell 2002).

In reference to genetic analysis in the Chilean scallop, few studies have been reported (von Brand-Skopnik & Ibarra-Humphries 2002). Karyotype analysis in A. purpuratus, carried out by Gajardo et al. (2002) found evidence of interpopulational chromosome differences, particularly for the short arm for chromosome pairs 2, 4, 6, 7, 8, 9, and 10 (2n = 32). The populations analyzed were two from commercial hatchery and one from the wild population. Unfortunately population genetic data are scarce and not yet available from genetic markers that substantiate the differences found in genome size. According to this scenario, it is difficult to attribute that the direction of genome size change would be due to domestication process. However, morphologic analysis showed significant differences between Tongoy and Arica height/length ratio (P < 0.0001). Future considerations about genome size together with molecular tools in both populations allow gathering deeper insight into the genetic characteristic of the northern Chilean scallop Argopecten purpuratus.

TABLE 1.
Genome size measurements obtained by fluorescence fading method
in two populations of the northern scallop A. purpuratus.

  Arica         Number
Population    of Nuclei    C-value    Std. Dev.    Std. Err.

    1            420        1.106      0.177        0.009
    2            168        1.023      0.094        0.011
    3            387        1.053      0.319        0.016
    4            116        1.063      0.130        0.012
    5            194        1.120      0.156        0.011
    6            601        1.088      0.093        0.004
    7            930        1.103      0.148        0.005
    8            595        1.043      0.249        0.010
    9            370        0.932      0.121        0.006
    10           457        1.094      0.150        0.007
    11           257        1.016      0.319        0.021
    12           379        0.994      0.259        0.013
    13           333        1.078      0.148        0.008
    14           223        1.179      0.314        0.021
    15           188        1.089      0.172        0.018
  Total         5,628       1.057 *    0.057 *      0.015 *

  Tongoy       Number
Population    of Nuclei    C-value    Std. Dev.    Std. Err.

    1            463        1.164      0.284        0.013
    2            446        1.084      0.181        0.009
    3            135        1.082      0.105        0.009
    4            427        1.201      0.389        0.019
    5            654        1.044      0.243        0.010
    6            485        1.197      0.173        0.008
    7            942        1.261      0.300        0.010
    8            396        1.106      0.171        0.009
    9            555        1.231      0.285        0.012
    10           450        1.198      0.267        0.013
    11           486        1.070      0.172        0.008
    12           502        1.285      0.397        0.018
    13           662        1.093      0.210        0.008
    14           515        1.141      0.216        0.010
    15           420        1.106      0.177        0.009
  Total         7,538       1.139 *    0.066 *      0.017 *

* Mean values.

TABLE 2.
ANOVA of genome size in populations of A. purpuratus.

Source of Variation       SS       df        ms       F       P

Population              15.87         1    15.87    277.3    ***
[Population.sup.*]      59.82        28     2.12     37.0    ***
  individual
Error                  739.67    12,926     0.06

*** P < 0.001.

ACKNOWLEDGMENTS

The authors thank Nicole Urriola for her technical assistance during sampling. This study was supported by the Mexican National Council of Science and Technology (Project 36075-B and Project 33018-B). The scallop samples were provided by Dr. Eduardo Uribe T. (FNDR project 20192982-0). The first author is a CONICYT-BID (Chile) Ph. D. fellow.

LITERATURE CITED

Bertino, B., W. Knape, M. Pytlinska, K. Strauss & J. Hammou. 1994. A comparative study of DNA content as measured by flow cytometry and image analysis in 1884 specimens. Anal. Cell. Pathol. 6:377-394.

Canello, F., L. Paredes & J. Toro. 1992. Induccion ala triploidia en el ostion del norte Argopecten purpuratus, por medio de shock termico de calor. Invest. Pesqueras 37:5-11.

Cavalier-Smith, T. & M. J. Beaton. 1999. The skeletal function of nongenic nuclear DNA: new evidence from ancient cell chimaeras. Genetica 106:3-13.

Dolezel, J., J. Bartos, H. Voglmayr & J. Greilhuber. 2003. Nuclear DNA content and genome size of trout and human. Cytometry 51A(2):127-128.

Gajardo, G. & J. Nunez. 1992. Seed production, the critical factor for mollusk farming in Chile. World Aquaculture 24:72-77.

Gajardo, G., M. Parraguez & N. Colihueque. 2002. Karyotype analysis and chromosome banding of the Chilean-Peruvian scallop Argopecten purpuratus (Lamarck, 1819). J. Shellfish Res. 21(2):585-590.

Gallardo-Escarate, C., J. Alvarez-Borrego, M. A. Del Rio-Portilla & V. Kober. 2004. A new method for genome size estimation by quantification of fluorescence fading. (Submitted).

Gregory, R. 2000. Nucleotypic effects without nuclei: genome size and erythrocyte size in mammals. Genome 43:895-901.

Gregory, R. 2001 a. Coincidence, coevolution, or causation? DNA content, cell size, and the C-value enigma. Biol. Rev. 76:65-101.

Gregory, R. 2001b. Animal genome size database. Available at: http:// www.genomesize.com.

Gregory, R. 2002. Genome size and developmental complexity. Genetica 115:131-146.

Gregory, R. 2003. Genome size estimates for two important freshwater mollusks, the zebra mussel (Dreissena polymorpha) and the schistosomiasis vector snail (Biomphalaria glabrata). Genome 46:841-844.

Gonzalez-Tizon, A. M., A. Martinez-Lage, I. Rego, J. Ausio & J. Mendez. 2000. DNA content, karyotypes, and chromosomal location of 18S-5.8S-28S ribosomal loci in some species of bivalve molluscs from the Pacific Canadian coast. Genome 43:1065-1072.

Hardie, D., R. Gregory & P. Hebert. 2002. From pixels to picograms: a beginners' guide to genome quantification by Feulgen image analysis densitometry. J. Histochem. Cytochem. 56(6):735-749.

Hinegardner, R. 1974. Cellular DNA content of the Mollusca. Comp. Biochem. Physiol. 47A:447-460.

Hirschfeld, T. 1979. Fluorescence background discrimination by prebleaching. J. Histochem. Cytochem. 27:96-101.

Johnson, G. D. & G. M. Nogueira-Araujo. 1981. A simple method of reducing the fading of inmunofluorescence during microscopy. J. Inmunol. Methods 43:349-350.

Johnson, G. D., R. Davidson, K. McNamee, G. Russell, D. Goodwin & E. Holborow. 1982. Fading of inmunofluorescence during microscopy: a study of the phenomenon and its remedy. J. Inmunol. Methods 55:231-242.

Kidwell, M. 2002. Transposable elements and evolution of genome size in eukaryotes. Genetica 115:49-63.

Komaru, A., Y. Uchimura, H. Ieyama & K. Wada. 1988. Detection of induced triploid scallop, Chlamys nobilis, by DNA microflurometry with DAPI staining. Aquaculture 69:201-209.

Martinez, G. & H. Perez. 2003. Effect of different temperature regimes on reproductive conditioning in the scallop Argopecten purpuratus. Aquaculture 228:153-167.

Navarro, J., G. Leiva, G. Martinez & C. Aguilera. 2000. Interactive effects of diet and temperature on the scope for growth of the scallop Argopecten purpuratus during reproductive conditioning. J. Exp. Mar. Biol. Ecol. 247:67-83.

Ono, M., T. Marakami, A. Kudo, M. Isshiki, H. Sawada & A. Segawa. 2001. Quantitative comparison of anti-fading mounting media for confocal laser scanning microscopy. J. Histochem. Cytochem. 49(3):305-311.

Orgel, L. & F. Crick. 1980. Selfish DNA: the ultimate parasite. Nature 284:604-607.

Rodriguez-Juiz, A. M., M. Torrado & J. Mendez. 1995. Genome-size variation in bivalve molluscs determined by flow cytometry. Mar. Biol. 126:489-497.

Thorgaard, G. H., P. S. Rabinovitch, M. W. Shen, G. A. Gall, J. Propp & F. M. Utter. 1982. Triploid rainbow trout identified by flow cytometry. Aquaculture 29:305-309.

Uozu, S., N. Ohmido, H. Ohtsubo, E. Ohtsubo & K. Fukui. 1997. Repetitive sequences: cause for variation in genome size and chromosome morphology in the genus Oryza. Plant Molecular Biology 35:791-799.

Vieira, C., C. Nardon, C. Arpin, D. Lepetit & C. Biemont. 2002. Evolution of genome size in Drosophila. Is the invader's genome being invaded by transposable elements? Mol. Biol. Evol. 19:1154-1161.

von Brand-Skopnik, E. & A. M. Ibarra-Humphries. 2002. Genetica de Pectinidos Iberoamericanos. In: A. N. Maeda-Martinez, editor. Los moluscos pectinidos de Iberoamerica: ciencia y acuicultura. Mexico: Noriega Editores. pp. 105-126.

von Brand, E. & A. Kijima. 1990. Comparison of genetic markers between the Chilean scallop Argopecten purpuratus and Japanese scallop Patinopecten yessoensis. Tohoku J. Agri. Res. 41(1-2):25-35.

von Brand, E., G. Bellolio & K. Lohrmann. 1990. Chromosome number of the Chilean scallop Argopecten purpuratus. Tohoku J. Agri. Res. 40(3-4):91-95.

Winkler, F., B. Ladron de Guevara, B. Estevez & L. Jollan. 1993. Induccion a la triploidia en Argopecten purpuratus Lamarck, 1819, mediante citocalasina-B. Rev. Biologia Marina 28:239-246.

CRISTIAN GALLARDO-ESCARATE, (1,2) JOSUE ALVAREZ-BORREGO, (2) ELISABETH VON BRAND-SKOPNIK (3) AND MIGUEL ANGEL DEL RIO-PORTILLA (1)

(1) Departamento de Acuicultura, Division de Oceanologia, (2) Direccion de Telematica, Centro de Investigacion Cientifica y de Educacion Superior de Ensenada, Km 107 Carretera Tijuana--Ensenada, Codigo Postal 22860, Ensenada, B.C. Mexico; (3) Departamento de Biologia Marina, Facultad de Ciencias del Mar, Universidad Catolica del Notre, Larrondo 1281, Casilla 117, Coquimbo, Chile

* Corresponding author. E-mail: [e-mail]