Filling in the gaps impeding the instigation of selective breeding programs in barramundi Lates calcarifer: fate of genetic diversity through to harvest, estimation of genetic parameters and early prediction of family growth based on cellular processes

Barramundi (Lates calcarifer), also known as Asian seabass, is an emergent aquaculture fish species possessing excellent farming attributes. As a result of its euryhaline life-history, culture of barramundi is extremely flexible and the species displays fast growth in freshwater, brackish, and/or marine environments, where it can be farmed in ponds, cages, or intensive recirculation tank-based systems. Barramundi has been farmed for over 40 years in tropical waters of the Indo-West Pacific region and has recently been introduced into Europe and North America due to increasing consumer demand and high market value, and global production is rapidly increasing. Whereas a wealth of knowledge on the biology and culture of the species have contributed to a developing industry, barramundi farmers still rely on unimproved stocks and often experience inconsistent production as a result of high variability in fish growth and survival between batches. However, for the barramundi industry to meet its full potential in the future it will need to rapidly move towards the instigation of well-informed genetic improvement programs. In order to assist with the design and implementation of a selective breeding program for barramundi, this research aimed to address key knowledge gaps related to the genetic diversity of farmed populations and to the underlying genetic basis of commercially important growth traits. In addition, this research investigated larval progeny testing as a means to rapidly estimate breeding values (EBV) of broodstock currently held at commercial hatcheries, or broodstock to be incorporated into a breeding program.\ud \ud One significant challenge for genetic studies of farmed barramundi relies on the fact that the species is a strict mass spawner and therefore the number and size of families cannot be fully manipulated by strip spawning and artificial fertilization. Such reproductive behavior has limited the knowledge on the genetic diversity of farmed cohorts, as pedigrees can only be reconstructed through DNA parentage analysis. Therefore, exploring how many families may be produced through mass spawning and, most importantly, retrieved at the end of the production cycle is critical for the development of breeding programs for barramundi. To investigate the fate of genetic diversity over the entire culture cycle, offspring from three commercially produced batches were genotyped with microsatellite markers at 18 days post hatch (dph) and then later at harvest (273-469 dph) for inference of familial relationships.\ud \ud Results demonstrated that it is possible to produce and retrieve at harvest over 100 families from a single spawning event using large numbers of broodstock (e.g. 33 fish). In addition, genetic diversity within generations was found much more resilient than previously thought. Despite significant changes in relative family frequencies over the culture period (P < 0.05), family lines identified at the larval stage were still available for selection by the time of harvest. The stability of several indices to gauge possible changes across the two sampling periods (i.e. effective population sizes, inbreeding rates, number of alleles, allelic richness, observed and expected heterozygosity and Rxy relatedness coefficient) also confirmed no subsequent loss of genetic diversity throughout the culture cycle. Therefore, early demographic census would provide an accurate representation of genetic diversity available for selection at harvest. Moreover, a cost-effective genotyping sample scheme targeting to reduce the overall costs of breeding programs of barramundi was also investigated. Here, examination of family contributions and mean family body weights at harvest showed that genotyping the top 1.5% of the population (or 750 individuals > 2.17 S.D. heavier than the population mean) will capture > 75% of family-specific genetic diversity present, whereas continued sampling after this was ineffective. This study suggests that genetic diversity in a barramundi breeding nucleus can be boosted and maintained by using several large breeding groups per generation.\ud \ud However, before efficient selective breeding programs can be implemented, it is also fundamental to have a comprehensive understanding of key genetic parameters of growthrelated traits, such as heritabilities (h²), genetic correlations (r(g)) and genotype by environment interactions (G x E). This knowledge is necessary to predict genetic gains and the efficiency of a breeding program, as well as to determine if genotypes selected under a particular environment will express the same growth performance when reared in a different environment. Heritability estimates for three separate batches measured and genotyped at harvest were consistently high for fish weight (mean W h² ~ 0.40), standard length (mean Ls h² ~ 0.37) and body depth (mean BD h² ~ 0.40). As additive genetic effects play a significant role in barramundi body size, selection of heavier (or longer/deeper) fish at harvest is expected to greatly improve fish growth rates. Lower heritability values were found for Fultons condition factor (0.00 ≤ K h² ≤ 0.20) and body shape (0.00 ≤ H h² ≤ 0.12), however, positive genetic correlations with fish weight (0.36 ≤ r(g) ≤ 0.41) indicate that selection for heavier fish may also improve fish condition and shape. Also of importance, very high genetic correlations were found for the same growth trait (W, Ls or BD) in fish reared in either fresh vs. sea water cages (r(g) ≥ 0.97), or commercially reared in an intensive tank system vs. a semi-intensive pond (r(g) ~ 0.99). The lack of G x E interactions suggest that the offspring of breeding candidates selected in a breeding nucleus, or in a particular farm, will express superior growth rates in a range of farming conditions which are representative for the barramundi industry. The lack of G x E interactions coupled with high heritability for growth are encouraging outcomes for the wider barramundi industry and the shared investment in a breeding program should return high genetic gains for all stakeholders alike.\ud \ud Nevertheless, the benefits from a future barramundi breeding program would only be capitalized after a few generations, i.e. a decade or longer. Therefore, alternative means to rapidly ascertain the genetic merit for growth rate of existing broodstock would be highly desirable. In this study, several larval progeny traits strongly associated with the individual growth capacity were investigated as potential early predictors of long-term family growth. High genetic correlations between early and late growth traits would indicate the feasibility of larval progeny testing as a method to infer EBVs of parental broodstock. Firstly, a protocol to measure larval size (Ls), total RNA, total DNA, total protein, RNA/DNA, protein/DNA and the proportion of cells dividing within a single larva was developed. Secondly, the heritability of these traits was estimated from a dataset of ca. 400 18 dph larvae with known pedigrees. Thirdly, the genetic correlations between heritable larval traits and fish harvest size (weight (W) and Ls)) were determined by combining the larval dataset with that of ca. 2000 harvested fish with known pedigrees. All larval traits exhibited moderate to high heritability at 18 dph (0.19 ≤ h² ≤ 0.51), indicating that their expression is under additive genetic control and therefore that they could have predictive power to estimate parental EBV for long term growth. This was confirmed by positive genetic correlation between all larval traits (except protein/DNA) and fish harvest W and Ls (r(g) ≥ 0.60). In particular, larval RNA/DNA, total RNA and the proportion of cells dividing had the highest predictive power to determine genetic differences in growth potential among barramundi broodstock shortly after spawning (0.81 ≤ r(g) ≤ 0.88). Results indicate that testing cellular traits in larval progeny may allow hatchery managers to immediately re-spawn the highest EBV ranked broodstock for stocking grow-out systems, thereby avoiding costs associated with rearing slow-growing families. Furthermore, early progeny testing may allow fish breeders working with long-lived and highly fecund multiple spawners like barramundi to cull introduction of broodstock with inferior genes for growth.\ud \ud In summary, this research bridges several missing links in the knowledge of genetic diversity and parameters of farmed barramundi necessary for the establishment of a selective breeding program for the species. It has been shown for the first time that despite the mass spawning nature of the species, sufficient genetic diversity can be generated and maintained in a breeding nucleus through to harvest and that growth rates can greatly be improved through selection for increased body weight at harvest. Furthermore, this research unveiled the existence of significant genetic correlations between cellular larval traits and fish harvest size which may be explored as a novel progeny testing tool to fast-track the improvement of fish growth.\ud \ud In summary, this research bridges several missing links in the knowledge of genetic diversity and parameters of farmed barramundi necessary for the establishment of a selective breeding program for the species. It has been shown for the first time that despite the mass spawning nature of the species, sufficient genetic diversity can be generated and maintained in a breeding nucleus through to harvest and that growth rates can greatly be improved through selection for increased body weight at harvest. Furthermore, this research unveiled the existence of significant genetic correlations between cellular larval traits and fish harvest size which may be explored as a novel progeny testing tool to fast-track the improvement of fish growth.\u