Selection response for growth rate (shell height and live weight) in the Chilean blue mussel (Mytilus chilensis Hupe 1854).

ABSTRACT The parental stock was taken from the 1999 natural spatfall of Mytilus chilensis collected in the Yaldad Bay of southern Chile. From the 22-mo-old cohort 5,688 mussels were monitored for live weight and shell length. Selection was carried out by applying a selection intensity of 1.755 for the trait "live weight". Five selected lines and five lines of an unselected control group were conditioned in seven 150-L tanks. Juveniles from the 3 selected and 2 control lines were individually tagged and transferred to three geographically distant mussel farms in southern Chile. Live weight and shell length were monitored after 10, 14, 18, and 22 month of age in all experimental mussels. The ANOVA results showed a significant difference, in both waits, between the selected and control groups at every age and location. Realized heritabilities for the trait "live weight" ranged between [h.sup.2] = 0.35 and [h.sup.2] = 0.54, whereas those for the trait "shell height" ranged between [h.sup.2] = 0.32 and [h.sup.2] = 0.49. Genotype-environment interactions were not apparent for either trait, indicating that similar selection pressures result in similar phenotypic changes for these traits across environments. These results suggest that mass selection for the improvement of the traits live weight and shell height would be effective in the Chilean mussel broodstocks.

KEY WORDS: mussel, Mytilus ehilensis, selection response, growth, Chile

INTRODUCTION

Mussel aquaculture has enormous potential in southern Chile mainly because the numerous sheltered bays and estuaries along its coastline are free from industrial pollution. Fertilizers and pesticides are rarely applied in local agriculture, and domestic pollution, because of the low human population densities in southern coastal areas. (Winter et al. 1984, Toro & Chaparro 1990).

The Chilean blue mussel Mytilus chilensis (Hupe 1854) is an economically important bivalve in southern Chile. Its culture began in 1943 in the area of Quellon located southeast of Chiloe Island (Navarro & Gutierrez 1990, Aiken 1993) and the aquaculture production of this species increased from 3,864 t in 1993 to 41,406 t in 2001 (SERNAPESCA 2002). These values indicate an increase of 971.5 % (37,542 t) for mussel aquaculture between 1993 and 2001. Greater production space (hectares) under cultivation rather than higher outputs per unit area accounts for most of this increased production (Sanchez 2002). However, social, ecological, and economical pressures will eventually promote a greater efficiency of these aquaculture production systems. There are various means to achieve this goal: improved husbandry, nutrition and disease control, and improvement by genetic means. The potential for improving performance by genetic means is unquestionably substantial. There are several reports in the literature addressing the importance and potential of genetics in aquaculture (Moav 1976, Newkirk 1983, Gjerde 1986, Guinez 1988, Toro & Newkirk 1990, Gjedrem 1999, Koment 2002, Beaumont & Hoare 2003). Increasing production efficiency, reducing generation time, and/or improving survival rate will reduce the negative effects of short growing seasons in high latitudes (Newkirk 1980).

Growth rate has implications for the fitness of individuals and also is considered the most economically important trait to be improved in a selection program (Beaumont & Hoare 2003). Traits of economic importance are usually determined by a large number of genes each having a small contribution and under considerable environmental influence (Newkirk 1980, Lande 1982). Thus, these traits are best studied through the quantitative genetic theory that reduces the most important features of complex genetic systems to a few variables that can be estimated from phenotypic measurements (Falconer 1989, Lande 1982).

Due to the high market demand for this species, the culture of M. chilensis in Chile has been increasing rapidly in the past years (Winter et al. 1984, Navarro & Gutierrez 1990, SERNAPESCA 2002, Sanchez 2002), and because of this mussel aquaculture development, estimates of genetic variation (i.e., heritability) and response to selection pressure are not only important from the ecological point of view but also for practical management purposes (Mallet et al. 1987, Koment 2002). The type of selection scheme that is appropriate depends upon the relative magnitude of the variation of that trait, due to differences in genotype and in environmental factors (Falconer 1989). These mussels are marketable at a shell length of about 55 mm, which is attained after 14 to 16 mo of growth in suspended culture (Winter et al. 1984, Kino & Valencia 1990). Therefore, a reduction in the time required to attain commercial size (12 mo) could be an important benefit in the culture of this mussel.

Very little has been published on the population genetics of M. chilensis and there are no reports on genetic improvement of the Chilean blue mussel. Control of the complete life cycle, which is required for any type of direct genetic improvement (Hershberger et al. 1984), has been accomplished only recently (Toro et al. 2004).

A high intrapopulation variation in the growth rate of M. chilensis has been reported in the literature (Winter et al. 1984, Navarro & Gutierrez 1990); however, only a few reports exist on genetic parameters (Toro & Paredes 1996) or genetic manipulation (Toro & Sastre 1995) in M. chilensis. A review of the literature shows that heritability estimates in blue mussels have been mainly based on half and/or full-sib correlation (Stromgren & Nielsen 1989, Toro et al. 2004), however, the most appropriate way in a commercial context to determine the heritability of a trait is to carry out selection trials and measure the response to selection for the trait (Beaumont & Hoare 2003). To ensure similar rearing conditions for the selected populations, a control group should also be spawned and their offspring reared in identical environments.

A high heritability value for a particular trait indicates that a large proportion of the trait is associated with additive gene action, and that genetic progress can therefore be achieved through mass selection (Newkirk et al. 1977). In this study, we consider that [h.sup.2] estimates using realized heritability analysis to obtain a more accurate estimation of its values among different rearing locations. To date, there have been no reports on growth and survival of hatchery produced seed of M. chilensis. This study represents the first attempt to produce large numbers of juveniles and adults under laboratory conditions and corresponds to a portion of a large-scale breeding program to improve the growth rate of the Chilean mussel Mytilus chilensis.

MATERIAL AND METHODS

Adult mussels from a 1999 naturally settled cohort (Mytilus chilensis Hupe 1854) were collected from a mussel farm located in Yaldad Bay, Chiloe Island (43[degrees]08'S; 73[degrees]44'W), southern Chile during October 2001 (prior to the spawning season). At 22 mo of age, a frequency distribution for the trait "live weight" from a random sample (N = 5,688) of the cohort of mussels was carried out. The trait presented a normal distribution (P > 0.05, Kolmorov Smirnov test, P > 0.05). A selection intensity of 1.755 was applied, corresponding to the largest 10% of the trait distribution. A similar number of average-size animals were segregated as control-line parents (Table 1).

Spawning was conducted over a 2-d period. The mussels were kept out of water for about 4-5 h before attempting spawning by thermal shock. They were rinsed in clean seawater and the three selected lines (N = 200 each) and 2 control lines (N = 200 each) were placed in ten 150-L tanks with filtered (1-[micro]m) and UV-treated sea water (FSW) at 18[degrees]C. They were then continuously monitored until they spawned at which time they were removed. The day after the spawning the water of each tank was filtered through a 45-[micro]m mesh. The embryos collected from each family were placed into a 200-L fiberglass tank containing 1-[micro]m filtered and UV-treated fresh seawater (FSW) at 16 [+ or -] 1[degrees]C, and a density of 100 individuals per mL.

After 48 h the density was adjusted to 5 larvae per mL. A high cell concentration (100,000 cells/mL) of the microalgae Isochrysis galbana (Parke 1949) and Chaetoceros gracilis (Schuett) were used as food (Toro & Paredes 1996). Every day the water in each tank was passed through a 45-[micro]m "nitex" screen to retain the larvae. Each beaker was rinsed with fresh water followed by seawater. The larvae were then resuspended in FSW and algal food was added daily at the desired cell concentration.

The juveniles from each selected and control lines were settled on netlon mesh and after 2 weeks of growth in the laboratory, transferred to the field (Quetalmahue) until they were ready for tagging (5 mm). After 15 weeks of growth in the field, the juveniles from each line were individually labeled and transferred randomly in pearl nets to 3 different aquaculture farming sites. Hueihue Bay, located 500 m from a commercial oyster farm; and Quetalmahue Gulf and Putemun Channel, located around a mussel farm, all of them within the Chiloe Island (Fig. 1).

Live weight and shell height were monitored at 10, 14, 18, and 22 mo of age using an A&D portable balance ([+ or -] 0.01 g) and vernier calipers ([+ or -] 0.1 mm).

Analysis of Variance (ANOVA) including the effects of location, group (selected and control), and their interactions were carried out for live weight and shell height at the four life stages using the SYSTAT 5.1 statistical package (Wilkinson 1991). Log-transformations were used to normalize the data where appropriate. The response to selection and realized heritability estimates and their standard errors were calculated according to Falconer (1989), Becker (1992), and Hadley et al. (1991). A statistical comparison (P [less than or equal to] 0.05) of differences in heritability values (Student's t test) was done.

RESULTS

Growth data at 10, 14, 18, and 22 mo of age for the traits live weight and shell height at 3 locations are presented in Table 2. From the mean of each group (control and selected) it can be seen that there are differences in both traits among the three different locations. The ANOVA results (Table 3) show that for live weight and shell height, there is a significant effect of "location" (three locations) and "group" (selected and control) sources of variance. There were no significant effects of the location by group interaction term. A significant response to selection for the direct selected trait "live weight" was found at all ages monitored (see Table 3). The standardized responses fluctuated between 0.77 and 0.05 units of SD from the control. A correlated response for the trait 'shell height' was detected (see Table 3). The response measured as standardized Index did not differ significantly among the different localities (P > 0.05).

The realized heritability estimates for the traits "live weight" and "shell height" at 22 months of age exhibited values significantly greater than zero in all three locations (Table 4). Heritabilities for live weight did not differ significantly between Hueihue and Quetalmahue locations (P > 0.05) while Putemtin and Quetalmahue showed similar heritabilities for shell height (P < 0.05); therefore even when growth rates differed significantly among environments, the additive genetic variance did not vary from location to location, with the exception of Putemtln and Hueihue for live weight and shell height respectively.

DISCUSSION

Live weight and shell height varied significantly among environments in which the offspring were raised (see Table 2 and 3). This implies that environmental variation plays a large role in determining these growth traits. The location x group interaction (G X E) was not significant for either of the two traits (see Table 3), meaning that although there was significant variation among environments for each trait, the overall rank of each group in each environment was unchanged across sites.

There was a significant and steady response to selection in both traits across the environments. These results are in accordance with those reported by Crenshaw et al. (1991), for Argopecten irradians concentricus; Bustos et al. (1991) and Toro et al (1994), for Ostrea chilensis; and Nell et al (2000), for Saccostrea glomerata were a positive response to selection was found. Toro and Newkirk (1991) working with O. chilensis reported also a significant response to reduce the same trait, comparing a low selected line against an unselected control line.

The realized heritability estimates in both traits 22 months of age (see Table 4), were significant across the three environments measured, although, no significant genotype x environment interaction for these traits where detected (see Table 3). Therefore, a selection intensity applied to these traits results in predictable changes in the phenotype values of the traits among environments (Falconer 1989). There are no reports on realized heritabilities on adult mussels, however, the heritability values obtained here are lower than those reported by Newkirk & Haley (1982), in Ostrea edulis; Lannan (1972), in Crassostrea gigas; Stromgren & Nielsen (1989) and Losee (1978), in Mytilus edulis. However, they are in accordance with those reported by Toro & Newkirk (1990), in O. edulis; Toro & Newkirk (1991), in O. chilensis; Wada (1986), in Pinctada fucata martensii ([h.sup.2] = 0.47 & 0.35 for shell width and shell convexity respectively); Hadley et al. (1991), in Mercenaria mercenaria ([h.sup.2] = 0.43 for growth rate); and also in line with values for production traits in livestock (Van Vleck 1987).

Using the realized heritability values obtained and applying a selection intensity of 1.755 that is equivalent to select the higher 10% of the population for the trait, the estimated mean change in live weight ranges between 2.6 % to 4.0 % and for shell height between 2.9 % to 4.5 % per generation. These values of response to selection for increasing the traits are lower than those given by Mallet et al (1986) and Stromgren & Nielsen (1989) for M. edulis and Newkirk & Haley (1982) for O. edulis; these authors reported a large response to selection for growth to market size with an average of 23 % of gain over the controls.

The correlated response found for the trait "shell height" suggests a positive genetic correlation between the two traits. Very high genetic correlations between live weight and shell height have been reported ([r.sub.a] = 0.995) for O. edulis (Toro & Newkirk 1990) and for M. edulis juveniles (Nielsen 1985).

The estimates of heritability for traits of economic importance such as live weight and shell height is essential in a long term breeding program to enhance the growth rate of the Chilean mussels. The quantitative genetic parameter estimates should be obtained under natural conditions and their constancy measured among different environments to be able to predict potential responses to selection (Mitchell-Olds & Rutledge 1986). Heritability values are essential for the purpose of predicting the genetic gain by selection. In common with previous studies of bivalves, our experiments demonstrate significant genetic variation in live weight and shell height, which may have implications for the biology of natural populations of M. chilensis as well as for commercial culture.

TABLE 1.

Mytilus chilensis. Basic statistics for the traits "live weight" (g)
and shell height (mm) for the 5.688 individuals from the 22 months
old cohort used as a base population in the selection program.

               N                    Population    Control    Selected
                                       5688         400        600
Live weight
  Mean (g)                            18.47        18.42      26.50
  Standard Deviation                   4.27         0.89       2.54
  Coefficient of Variation (%)        23.11         4.83       9.50
Shell height
  Mean (mm)                           63.05        63.41      69.73
  Standard Deviation                   5.24         3.11       3.95
  Coefficient of Variation (%)         8.31         4.90       5.66

TABLE 2.

Mytilus chilensis. Basic statistics for the traits "shell height" and
"live weight" in the F1 of the selected and control groups at 10, 14,
18, and 22 months old maintained in the 3 mussel farms: Hueihue,
Putemun, and Quetalmahue, (mean: average; SD: standard deviation).

                           Age       Live Weight     Shell Height
   Farm        Group     (months)        (g)             (mm)

Hueihue                              Mean     SD     Mean      SD
              Control       10       5.35    2.22   36.73     5.39
                            14       9.00    3.25   44.68     2.28
                            18      11.81    3.10   49.28     4.55
                            22      19.20    5.29   56.37     4.35
              Selected      10       6.55    1.51   40.89     4.50
                            14      10.36    2.36   48.23     4.30
                            18      12.30    2.69   52.78     4.15
                            22      22.34    4.18   58.42     4.28
Putemun                              Mean     SD     Mean      SD
              Control       10       5.53    2.35   37.60     5.87
                            14       9.02    2.89   45.12     5.13
                            18      12.10    3.01   49.90     4.37
                            22      18.77    6.16   56.50     5.47
              Selected      10       6.07    1.88   39.76     5.10
                            14       9.15    2.40   46.03     5.18
                            18      12.41    3.17   50.72     4.81
                            22      22.80    4.60   59.93     5.75
Quetalmahue                          Mean     SD     Mean      SD
              Control       10       7.58    2.52   42.68     5.39
                            14      12.38    3.39   50.27     5.31
                            18      15.20    3.62   54.20     4.76
                            22      21.52    4.48   58.24     5.04
              Selected      10       8.14    2.11   44.99     4.50
                            14      13.50    2.81   52.50     4.54
                            18      16.20    3.37   55.90     4.55
                            22      24.38    4.09   61.28     4.37

TABLE 3.

Mytilus chilensis. Analysis of variance (ANOVA) for the traits "shell
height" and "live weight" at 10, 14, 18, and 22 months old.

                             Shell Height       Live Weight

      Effect          df       MS       P        MS       P

10 Months
  Location              2   4586.79   <0.05    673.52   <0.05
  Group                 1    368.00   <0.05     22.10   <0.05
  Location * Group      2     75.93   >0.05      8.47   >0.05
  Error              1507     29.33              5.32
14 Months
  Location              2   2727.43   <0.05   1047.90   <0.05
  Group                 1    212.36   <0.05     54.22   <0.05
  Location * Group      2     66.05   >0.05     14.71   >0.05
  Error              1016     26.66              9.70
18 Months
  Location              2   1552.06   <0.05     11.34   <0.05
  Group                 1     96.84   <0.05     40.85   <0.05
  Location * Group      2     46.96   >0.05      7.44   >0.05
  Error               719     20.81             10.52
22 Months
  Location              2    348.51   <0.05    382.84   <0.05
  Group                 1   1465.70   <0.05   2030.38   <0.05
  Location * Group      2     30.38   >0.05     22.92   >0.05
  Error               747     23.75             23.26

df, degrees of freedom; MS, mean squares.

TABLE 4.

Mytilus chilensis. Realized heritability estimates ([h.sup.2]) and
their standard errors ([+ or -]SE) for the traits "shell height" and
"live weight" at 22 months old grown at 3 geographically separated
mussel farms.

                  Shell Height            Live Weight
                  [h.sup.2] SE           [h.sup.2] SE

Hueihue        0.32 [+ or -] 0.017    0.38 [+ or -] 0.020
Putemun        0.49 [+ or -] 0.024    0.54 [+ or -] 0.018
Quetalmahue    0.48 [+ or -] 0.016    0.35 [+ or -] 0.015

ACKNOWLEDGMENTS

We thank Instituto Fomento Pesquero, Centro Hueihue, for all the laboratory facilities to carry out this project.

LITERATURE CITED

Aiken, D. 1993. Bivalve culture in Chile. World Aquaculture 24:7-19.

Beaumont, A. R. & K. Hoare. 2003. Biotechnology and Genetics in Fisheries and Aquaculture, 1st ed. Oxford: Blackwell Science Ltd.

Becker, W. A. 1992. Manual of quantitative genetics, 5th ed. Pullman, Washington: Academic Enterprises.

Bustos, E., E. Pacheco & F. Ponce. 1991. Pesquerias bent6nicas chilenas. Rev. Pacifico Sur 19:1-13.

Crenshaw, J. W., P. B. Heffernan & R. L. Walker. 1991. Heritability of growth rate in the southern bay scallop, Argopecten irradians concentricus (Say, 1822). J. Shellfish Res. 10:55-63.

Falconer, D. S. 1989. Introduction to Quantitative Genetics. New York: Longman Group Limited.

Gjerde, B. 1986. Growth and reproduction in fish and shellfish. Aquaculture 57:37-55.

Gjedrem, T. 1999. Aquaculture needs genetically improved animals. The Advocate, Global Aquaculture Alliance 2:69-70.

Guinez, R. 1988. Mejoramiento genetico en recursos marinos: situacion actual y perspectivas, lnvestigacion Pesquera Chile 35:113-121.

Hadley, N. H., R. T. Dillon & J. J. Manzi. 1991. Realized heritability of growth rate in the hard clam Mercenaria mercenaria. Aquaculture 93:109-119.

Hershberger, W. K., J. A. Perdue & J. H. Beattie. 1984. Genetic selection and systematic breeding in Pacific Oyster Culture. Aquaculture 39: 237-245.

Kino, S. & J. Valencia. 1990. El informe final de las investigaciones de la repoblacion y cultivo de la ostra chilena y el erizo en la Isla de Chiloe en la Reptiblica de Chile. Overseas Fishery Cooperation Foundation. Informe Tecnico.

Koment, H. 2002. Defining breeding goals for future sustainable aquaculture. World Aquaculture 33:11-14.

Lande, R. 1982. A quantitative genetic theory of life history evolution. Ecology 63:607-615.

Lannan, J. E. 1972. Estimating heritability and predicting response to selection for the Pacific oyster, Crassostrea gigas. Proc. Natl. Shellfish Assoc. 62:62-66.

Losee, E. 1978. Influence of heredity on larval and spat growth in Crassostrea virginica. Proc. World Maricult. Soc. 9:101-107.

Mallet, A. L., K. R. Freeman & L. M. Dickie. 1986. The genetics of production characters in the blue mussel Mytilus edulis. A preliminary analysis. Aquaculture 57:133-140.

Mallet, A. L., C. Carver, S. Coffen & K. Freeman. 1987. Winter growth of the blue mussel Mytilus edulis: importance of stock and site. J. Exp. Mar. Biol. Eco1.108:217-228.

Mitchell-Olds, T. & J. J. Rutledge. 1986. Quantitative genetics in natural populations: a review of the theory. Am. Nat. 127:379-402.

Moav, R. 1976. Genetic improvement in aquaculture industry. In: T. V. R. Pillay W. A. Dill, editors. Advances in aquaculture. Surrey, England: Fishing News Books Ltd. pp. 610-622.

Navarro, J. M. & A. Gutierrez. 1990. Sinopsis sobre el estado actual del cultivo de mitilidos en Chile. In: A. Hernandez, editor. Cultivo de moluscos en America Latina. Bogota. pp. 189-210.

Nell, J. A., I. R. Smith & C. C. McPhee. 2000. The Sydney rock oyster Saccostrea glomerata (Gould 1850) breeding programme: progress and goals. Aquac. Res. 31:45-49.

Newkirk, G. F. 1980. Review of the genetics and the potential for selective breeding of commercially important bivalves. Aquaculture 19:209-228.

Newkirk, G. F. 1983. Applied breeding of commercially important molluscs: a summary of discussion. Aquaculture 33:415-422.

Newkirk, G. F. & L. E. Haley. 1982. Progress in selection for growth rate in the European oyster Ostrea edulis. Mar. Ecol. Prog. Ser. 10:77-79.

Newkirk, G. F., L. E. Haley, D. L. Waugh & R. Doyle. 1977. Genetics of larvae and spat growth rate in the oyster Crassostrea virginica. Mar. Biol. 41:49-52.

Nielsen, M. V. 1985. Increase in shell length as a measure of production and ingestion of Mytilus edulis L. J. Exp. Mar. Biol. Ecol. 88:101-108.

Sanchez, V. 2002. Crece la industria mitilicultora. Aquanoticias 75:8-14.

SERNAPESCA. 2002. Anuario estadfstico de pesca, 2001. Servicio Nacional de Pesca, Ministerio de Economfa Fomento y Reconstruccion, Santiago, Chile.

Stromgren, T. & M. V. Nielsen. 1989. Heritability of growth in larvae and juveniles of Mytilus edulis. Aquaculture 80:1-6.

Toro, J. E. & G. F. Newkirk. 1990. Divergent selection for growth rate in the European oyster (Ostrea edulis L.): response to selection and estimates of genetic parameters. Mar. Ecol. Prog. Ser. 62:219-228.

Toro, J. E. & O. R. Chaparro. 1990. Conocimiento biol6gico de Ostrea chilensis (Philippi, 1845): Impacto y perspectivas en el desarrollo de la ostricultura en Chile. In: A. Hernandez, editor. Cultivo de moluscos en America Latina. Bogotg. pp. 231-264.

Toro, J. E. & G. F. Newkirk. 1991. Response to artificial selection and realized heritability estimate for shell height in the Chilean oyster Ostrea chilensis. Aquat Living. Resour. 4:101-108.

Toro, J. E. & D. Sastre. 1995. Induced triploidy in the Chilean blue mussel, Mytilus chilensis (Hupe 1854), and performance of triploid larvae. J. Shellfish Res. 14:161-164.

Toro, J. E. & L. Paredes. 1996. Feeding trial with different families of Chilean blue mussel, Mytilus chilensis larvae: evidence for genotype-food interaction. J. Appl. Aquac. 6:81-87.

Toro, J. E., P. Aguila, A. M. Vergara & G. F. Newkirk: 1994. Realized heritability estimates for growth from data on tagged Chilean native oyster (Ostrea chilensis). World Aquaculture 25:29-30.

Toro, J. E., A. C. Alcapan, A. M. Vergara & J. A. Ojeda. 2004. Heritability estimates of larval and spat shell height in the Chilean blue mussel (Mytilus chilensis Hupe 1854) produced under controlled laboratory conditions. Aquaculture Research 35:56-61.

Van Vleck, L. D. 1987. Observations in advances in dairy cattle. In: B. S. Weir, E. J. Eisen, M. M. Goodman, G. Namkoong, editors. Proceedings of 2nd Conference on Quantitative Genetics. Massachusetts: Sinauer Associates.

Wada, K. T. 1986. Genetic selection for shell traits in the Japanese pearl oyster, Pinctada fucata martinensii. Aquaculture 57:171-176.

Wilkinson, L. 1991. SYSTAT The system for statistics. Evanston, Illinois: Systat, Inc.

Winter, J. E., J. E. Toro, J. M. Navarro, G. S. Valenzuela & O. R. Chaparro. 1984. Recent developments, status and prospects of molluscan aquaculture on the Pacific coast of South America. Aquaculture 39:95-134.

This research was supported by Fondo Nacional de Investigacion Cientifica y Tecnologica, Chile. Project FONDECYT 1010166.

JORGE E. TORO, * ANGELICA C. ALCAPAN, JOHANA A. OJEDA AND ANA M. VERGARA

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

Instituto de Biologia Marina, Universidad Austral de Chile, Casilla 567, Valdivia, Chile.