Study on heritability of growth in the juvenile sea urchin Strongylocentrotus nudus

Journal of Shellfisheries Research, August, 2004 by Liu Xiaolin, Chang Yaqing, Xiang Jianhai, Ding Jun, Cao Xuebin

ABSTRACT The heritability of growth of juvenile Strongylocentrotus nudus was analyzed using quantitative genetic methods. Twenty-one half-sib groups and 60 to 63 full-sib groups of juveniles were obtained by artificial fertilization of three to five females by single males based on a nested design. The body weight (g) and test diameter (cm) of the young was measured 3 and 5 months after metamorphosis. Maternal component estimates are significantly greater than paternal component estimates for both weight and diameter at both ages. Greater maternal components suggest large non-additive genetic effects that could not be differentiated with the available data. Estimates of heritability in the narrow sense calculated from the additive genetic component using a paternal half-sib correlation analysis ranged from 0.2167-0.4565 for weight and 0.2059-0.4998 for diameter. The results indicate significant maternal effects. The strength of the nested design and the paternal half-sib correlation analysis used in this study make the estimate the most precise and unbiased reported to date.

KEY WORDS: sea urchin. Strongylocentrotus nudus, growth, heritability

INTRODUCTION

Sea urchins are one of the most important aquaculture species in the world. The gonads have long been used as a luxury food and as a food source by common people in many countries (Hobson & Chave 1990, Shimabukuro 1991, Hagen 1996). Because of over-fishing, interest in aquaculture of sea urchins has increased greatly (Hagen 1998, Lawrence et al. 2001). One of the most important species for aquaculture is Strongylocentrotus nudus (Hagen 1996, Agatsuma 1998). Current culture of S. nudus has used seeds obtained from wild individuals (Gao & Chang 1999, Liao & Qiu, 1999). Analysis of populations of S. nudus in Japanese waters shows considerable range in size of small individuals, presumably of single cohorts (Agatsuma 1997). This could result from inter action of genetic characteristics and the environment. Vadas et al. (2002) found evidence for intrinsic variability in field populations of Strongylocentrotus droebachiensis. It is important to document the degree of heritability of growth in sea urchins because of its implications for both fisheries and aquaculture.

Demonstration of heritability is best done with comparisons of half-sib groups because they are less likely to be affected by environmental influence (Gjedrem 1992). Sib analysis techniques have been used for several important aquaculture species (Mallet et al. 1986, Rawson & Hilbish 1990, Hadley et al. 1991, Crenshaw et al. 1991, Newkirk et al. 1977, Benzie et al. 1997). The purpose of this study is to estimate heritability of growth in terms of body weight and diameter of juvenile Strongylocentrotus nudus.

MATERIALS AND METHODS

Experimental Design

This study used a classic nested mating design developed by Comstock and Robinson (1952) to partition the phenotypic variation in juvenile growth into its genetic and non-genetic causes. In this experiment each of 21 male Strongylocentrotus nudus was mated to 3 to 5 females, therefore generating 63 full-sib families and 21 half-sib families. The effects of males and females nested within males on growth were separated using nested analysis of variance (ANOVA). Juveniles were weighed and their diameters measured at 3 and 5 months of age.

Genetic Analysis

The covariance among full- and half-sibs provides the basis for the separation of phenotypic variance into genetic and environmental components of variance. The covariance among full- and half-sibs are calculated from the observed components of variance obtained from a three-level nested, unbalanced ANOVA (Table 1) and the General Linear Models procedure of the statistical analysis system (SAS) (Freund et al. 1986).

The experiment was a three-level classic nested, unbalanced design. Therefore the number of offspring in dams and in sires and in dams within sires should revise ("revise" means adjust). The effective means were computed using the equations:

Effective mean number of offspring in dams within sires

[K.sub.1] = [N - [summation of]([n.sub.ij.sup.2]/[dn.sub.i])]/(D - S)

Effective mean number of offspring in dams:

[K.sub.2] = [[summation of]([n.sub.ij.sup.2]/[dn.sub.i]) - [summation of] ([n.sub.ij.sup.2]/N)]/(S - 1)

Effective mean number of offspring in sires:

[K.sub.3] = (N - [summation of][dn.sup.2.sub.i]/N)/(S - 1)

in which S = number of sires, D = number of dams, [n.sub.ij] = number of offspring of the i-th sire and j-th dam, [dn.sub.1] = number of offspring of i-th sire, N = sum of number of offspring of all sires or all dams.

The phenotypic variance ([V.sub.P]) was separated into the additive genetic variance ([V.sub.A]), non-additive genetic variance ([V.sub.N] and environmental variance ([V.sub.E]), and the environmental variance ([V.sub.E]) was separated into the common environmental variance ([V.sub.E]) and the specific environmental variance ([V.sub.ES) using the standard separation of variance components (Falconer 1989). The causal components of variance were estimated from the full- and half-sib covariance using the relationships in Table 2.