Elsevier

Aquaculture

Volume 318, Issues 3–4, 8 August 2011, Pages 325-334
Aquaculture

A novel breeding programme for improved growth in barramundi Lates calcarifer (Bloch) using foundation stock from progeny-tested parents

https://doi.org/10.1016/j.aquaculture.2011.05.037Get rights and content

Abstract

Rapid genetic gains for growth in barramundi (Lates calcarifer) appear achievable by starting a breeding programme using foundation stock from progeny tested broodstock. The potential gains of this novel breeding design were investigated using biologically feasible scenarios tested with computer simulation models. The design involves the production of a large number of full-sib families using artificial mating which are compared in common growout conditions. The estimated breeding values of their paternal parents are calculated using a binomial probit analysis to assess their suitability as foundation broodstock. The programme can theoretically yield faster rates of genetic gain compared to other breeding programmes for aquaculture species. Assuming a heritability of 0.25 for growth, foundation broodstock evaluated in two years had breeding values for faster growth ranging from 21% to 51% depending on the genetic diversity of stock under evaluation. As a comparison it will take between nine and twenty-two years to identify broodstock with similar breeding values in a contemporary barramundi breeding programme.

Introduction

Barramundi (Lates calcarifer) also known as Asian sea-bass is an increasingly important tropical aquaculture species of the Asia-Pacific region and it is inevitable that breeding programmes for this species will soon commence (Macbeth et al., 2002, Wang et al., 2008). We are not aware of any published papers showing genetic gains for barramundi, and know of only one simulated breeding programme recently reported (Robinson et al., 2010). At the onset of any new breeding programme in aquaculture there is much to be gained by assessing wild genetic diversity as different strains may be more suitable for commercial production. The walk-back selection programme for growth rate proposed by Robinson et al. (2010) does not attempt to evaluate the potentially diverse strains from different geographic locations prior to breeding. In species other than barramundi regional sampling of strains has revealed a 52% difference between low and high growth in six strains of Labeo rohita (Reddy et al., 2002), a 73% difference in weight in five strains of Onorhynchus mykiss (Overturf et al., 2003) and a 104% difference in weight at 105 days between Abbassa and Maryout tilapia strains (Elghobashy, 2001). Differences within lines can also be large with Brody et al. (1976) reporting differences between the means of half-sibs as large as 30% of the overall mean in Cyprinus carpio.

If the breeding values of wild fish from different regions could be evaluated prior to establishing a breeding programme then there is the potential to make significant genetic gains. Common practise in barramundi hatcheries is to source replacement broodstock from the wild when required, but some hatcheries are starting to use selected commercially grown fish. As with other aquaculture species a breeding programme is usually initiated with one or perhaps combined strains randomly sampled as foundation parents. To address the uncertainty in strain selection a two stage selection approach has been applied in the past where strains are previously evaluated (Elghobashy, 2001) prior to selecting the best strains for a foundation population. However, this strategy can be costly and can take considerably more time than simply forming a synthetic line of mixed strains. More recently in barramundi there have been attempts to find genetic markers linked to quantitative trait loci (QTL) of economic importance as a potential means of screening foundation broodstock (Wang et al., 2007). However, again this method is costly and is restricted to a small number of QTL with large effects so ignores the potentially largest component of genetic variance from cumulative effects of many genes with smaller effects.

In an alternative strategy the high accuracy of progeny testing (Robertson, 1957) could be used to evaluate wild fish. This strategy has been under consideration for many years since Wohlfarth et al. (1961) used it to assess growth in carp. Later Brody et al. (1976) advocated large scale progeny tests but Gjedrem (1983) suggested that it would “increase generation interval markedly”. Five years later Gall (1988) mentioned that there was no evidence that progeny testing had been successfully implemented in fish breeding and since then it has received little attention in aquaculture for testing of quantitative traits such as growth rate.

Barramundi is ideally suited to progeny testing because their high fecundity in both females (up to 46 × 106 eggs per female; Davis, 1984) and males (up to 10–15 ml of semen; Maneewong, 1986, Palmer, 2000) allows many progeny to be tested for each parent. Artificial fertilisation would be essential because large numbers of synchronous natural spawns are difficult to achieve in practise for this species. Artificial fertilisation can also eliminate maternal effects and eliminate age differences which could potentially give fish a size advantage they never relinquish (Tave, 1995). We propose screening potential foundation broodstock for growth using genotype identification and phenotypic observations in a progeny test framework where families are produced by artificial fertilisation.

While copious quantities of semen can be collected from wild males captured on spawning grounds, this is generally only possible a short time before spawning in captive males (Hogan et al., 1987). The potential to strip-spawn eggs and artificially inseminate them with cryopreserved semen from multiple sires has been successfully demonstrated in L. calcarifer (Palmer et al., 1993) and enables the progeny of many half-sib families to be grown for accurate breeding value determination of sires. The protandrous sex reversal of L. calcarifer (male to female at 3–8 years of age: Davis, 1982, Moore, 1979) offers a novel approach in which wild broodstock females can be accurately evaluated prior to selecting them as foundation parents from progeny testing of their paternal full-sib families. The breeding values of young males can also be determined with relatively high precision by combining information from their own phenotype with the relatively accurate breeding values of their progeny tested sires. Thus, young males can also be evaluated as possible foundation broodstock providing inbreeding is managed.

In general, to manage inbreeding to perhaps less than 1% per generation (Goddard, 1992, Meuwissen and Woolliams, 1994) many more broodstock are needed for a selective breeding programme compared to the relatively low numbers of broodstock that are needed solely to produce fingerlings for industry. This has perhaps been the most important factor that has hitherto inhibited the establishment of a barramundi breeding programme in Australia. Thus in designing a suitable programme for selective breeding in barramundi, as with other large aquaculture species, it is important to consider minimising broodstock numbers to manage costs while having sufficient numbers to manage inbreeding.

Minimising broodstock numbers is one method of reducing costs but what is perhaps more important is to maximise early genetic gains (Smith, 1978). We explore an option to improve the rate of early genetic gains using a mating plan with intense between-family selection of potential foundation stock accurately identified from progeny testing wild barramundi.

Stochastic computer methods are used to evaluate the progeny test scheme proposed here under a range of simulated parameter values. We examine how a progeny test scheme could be implemented for barramundi to estimate heritability, to assess geographic strains, and to achieve rapid genetic gains while managing inbreeding for long term selection. To assist the successful implementation of the scheme a description of husbandry methods is also presented in detail.

Section snippets

Methods

This is a computer simulation study with the general breeding design having five basic stages: (i) collect wild males and their milt for use in the progeny test, (ii) evaluate wild broodstock through a progeny test, (iii) select the very best progeny test sires (which change to females) and the very best young males from the progeny test as foundation stock, (iv) multiply the best foundation stock to create sufficient families to manage long term inbreeding and (v) perform ongoing selection in

Heritability

Heritability (h2) estimates using a binomial probit analysis from genotyping the heaviest (largest) NG progeny were consistent with the simulated heritability (Table 1). This confirms that the probit analysis is a suitable way of determining heritability for continuous traits such as harvest weight and that the simulation is implemented correctly.

Genetic gains

The progeny test was used to evaluate potential foundation stock prior to the commencement of a breeding programme. We first determine if a variation

Discussion

There is nothing new about progeny testing in aquaculture (Wohlfarth et al., 1961) or in using artificial mating to estimate genetic parameters (Dupont-Nivet et al., 2008). What is novel in this study is the way we combine strain evaluation, genetic parameter estimation and progeny testing to evaluate potential foundation broodstock using a carefully controlled mating design.

The necessity to regularly grade barramundi until they reach a size of about 250 g makes traditional genetic parameter

Conclusion

The L. calcarifer progeny test design described in this paper theoretically identifies superior foundation broodstock in two years. The same level of genetic improvement could take nine to 22 years of selection in other proposed barramundi breeding programmes. Despite the extra effort involved in sourcing potential foundation stock for the progeny test, and the challenging nature of the husbandry activities that are required, our scheme compares favourably with the risks that accrue over a much

Acknowledgements

The progeny test approach was inspired from discussions with Dr Roger Lewer while he was an employee of the Queensland Department of Primary Industries and Fisheries. Thanks to the many useful comments from referees and reviewers. Funding support was provided by the Department of Employment, Economic Development and Innovation.

References (44)

  • V. Parameswaran et al.

    A fish nodavirus associated with mass mortality in hatchery-reared Asian Sea bass, Lates calcarifer

    Aquaculture

    (2008)
  • P.V.G.K. Reddy et al.

    Growth and survival of six stocks of rohu (Labeo rohita, Hamilton) in mono and polyculture production systems

    Aquaculture

    (2002)
  • G.H. Yue et al.

    Genetic variation and population structure of Asian seabass (Lates calcarifer) in the Asia-Pacific region

    Aquaculture

    (2009)
  • S.F. Chenoweth et al.

    Concordance between dispersal and mitochondrial gene flow: isolation by distance in a tropical teleost, Lates calcarifer (Australian barramundi)

    Heredity

    (1998)
  • T.L.O. Davis

    Maturity and sexuality in barramundi, Lates calcarifer (Bloch), in the Northern Territory and South-eastern Gulf of Carpentaria

    Australian Journal of Marine Freshwater Research

    (1982)
  • T.L.O. Davis

    Estimation of fecundity in barramundi, Lates calcarifer (Bloch), using an automatic particle counter

    Australian Journal of Marine Freshwater Research

    (1984)
  • H. Elghobashy

    Aquaculture genetics research in Egypt. p. 29–34

  • D.S. Falconer

    Introduction to quantitative genetics

    (1972)
  • L.Ma.B. Garcia

    Spawning response of mature female sea bass, Lates calcarifer (Bloch), to a single injection of luteinizing hormone-releasing hormone analogue: effect of dose and initial oocyte size

    Journal of Applied Ichthyology

    (1989)
  • R.N. Garrett et al.

    Induced breeding of barramundi

    Austasia Aquaculture

    (1991)
  • D. Gianola et al.

    Sire evaluation for ordered categorical data with a threshold model

    Genetic Selection Evolution

    (1983)
  • A.R. Gilmour et al.

    ASREML User's Manual

    (2001)
  • Cited by (0)

    © The State of Queensland, Department of Employment, Economic Development and Innovation, 2010.

    View full text