Environment, genes and the effects of genotype by environment (G x E) interactions, on the expression of commercially important traits in the silver-lipped pearl oyster, Pinctada maxima (Jameson)

The silver-lipped pearl oyster, Pinctada maxima, is cultivated in Australia and throughout South East Asia for the production of "South Sea pearls". Pearls from P. maxima are the largest and most valuable pearl produced out of all pearl producing oyster species, with the industry being valued at ~US$ 412 million farm-gate globally. The commercial production of pearls from this species is well established and hatchery techniques have been developed to the stage where the life cycle is considered closed. A standard commercial pearl production cycle takes 4 years from the time oysters are spawned in the hatchery to the harvesting of pearls. The first two years is a "prepearl" grow-out phase for the oyster that will become the host oyster, and then a further two years in a "pearl growth" phase whereby the oyster is "seeded" and the pearl nacre is laid down. After this 4 year cycle only ~20% of pearls harvested fall within the category of being "gem quality", where the combination of the commercially important traits lustre, size, shape, weight, colour and complexion are all within the accepted standards. It is this long phase of production and the low proportion of gem quality pearls harvested that makes it very attractive for pearling companies to improve the consistency of production through the use of genetic breeding approaches. A current impediment to adoption of genetic approaches, however, is a lack of knowledge on the role additive genetic factors play in the realization of a pearl quality trait. Another consideration for future breeding programs is that pearl companies generally have several farm sites situated in geographically distinct locations, with one hatchery supplying all the seedstock to these locations. As such, as well as understanding the genetic basis underpinning pearl traits, it is important for companies to determine the effect disparate environmental influences due to site location may have on both oyster growth and pearl quality, and to establish whether the realization of genetic potential in improved oysters will be affected by genotype by environment (G x E) interactions. Like that for genetic parameters governing pearl quality and growth traits, however, data on the potential impact G x E may have on selected oysters reared under different environmental conditions is lacking. In response to this critical missing information this thesis aimed to establish genetic parameters and estimate genotype by environment interactions for both growth traits and pearl quality in the silver-lipped pearl oyster. Investigations commenced in this thesis by considering the effect long-line location (and perceived differences in micro-environment) at four sites within a pearl farm have on the realization of growth and pearl quality. Variance of these traits due to genetic differences as a consequence of using different cohorts of broodstock was also considered. In this experiment long-line site was shown to have a significant impact on the overall growth rate of oysters, with oysters reared on long-lines at the Sasanaflapo site growing significantly faster than those at the other three sites examined. These growth differences were also shown to manifest regardless of the age or genetic composition of the cohort evaluated. This shows that oyster growth is a trait that may be influenced by environmental parameters within farm locations, and that long-line location may override individual genetic effects on growth. The influence site has within a farm on pearl quality was, however, less conclusive. Only differences in pearl colour overtones and lustre appeared to be modified by site effects within a farm. For example, pearl colour and lustre could be both partitioned using classification tree analyses by site, with the Duyef and Wulu sites producing on average more white pearls with pink overtones, while Maratlap and Sasanaflapo produced higher numbers of silver pearls with pink overtones. Likewise for lustre, splits in classification trees were related to the fact that the Wulu site produced more pearls exhibiting higher lustre grades. Despite these differences though no disparities were found in the economic value of pearls harvested from the various sites within the farm evaluated. With the suggestion that local-scale environmental effects may modify oyster growth traits the thesis then goes on to test if genetic differences represented by oysters from different families could be easily modified through the manipulation of defined environmental parameters. Here the relative performance in shell growth of spat from five full-sib families when spat were communally reared at different salinities (29, 34 and 40 ppt), food availability (high, medium and low), food quality (high, medium and low), and in a hatchery vs. ocean environment for 43 days, were compared. In support of the first experiment, rearing environment was again found to significantly influence growth expression, with significant differences evident when spat were reared at different salinities, in the ocean instead of hatchery, or when fed algae of differing nutritional quality. Additionally, comparative family growth was also altered when the environment changed, with significant environment by family interactions (G x E) apparent under food quality, food availability and hatchery vs. ocean rearing conditions. These results indicate, that at least during early oyster growth phases, that growth and relative family performance in P. maxima may change dependent on local environmental conditions. To further examine the effect environment and genetics has on oyster growth, spat were produced for a large commercial scale trial using broodstock originating from three distinct Indonesian populations (Bali, West Papua, Aru Islands). These spat from different genetic backgrounds were communally on-grown to adult sizes for two years at each of two Indonesian commercial farming locations (Bali and Lombok). Microsatellite based parentage determination analyses were used to retrospectively sort out oysters to both their family and population of origin. Significant size differences were observed in all shell growth traits measured (dorsal-ventral measurement DVM, anterior-posterior measurement APM, shell width SW and wet weight WW), with oysters originating from Bali and West Papua (DVM (mm) = 103.7±0.9 and 101.0±0.6 respectively) growing faster than those from the Aru population (93.5±0.5) at both grow-out locations. Family level differences within these populations were also present for shell traits, indicating a large amount of genetic variability present for potential breeding programs. However, although there were significant familial size differences for shell traits, unlike the earlier spat growth experiment, genetic correlation analyses showed little evidence for re-ranking of family performance among the two culture sites (r(g) = 0.89–0.99). This implies that under the commercial conditions oysters were evaluated that insignificant genotype by environment deviations among sites were evident. Heritability analyses based on these families were also conducted for shell traits, with DVM and APM found to be moderately heritable (0.15 ±0.00 (DVM), 0.23 ±0.03 (APM)). Thus selection for faster growing host oysters should be possible which would advantage the industry by decreasing the amount of time it takes to grow oysters to a size suitable for nucleus implantation. The final experiment undertaken in this thesis estimated for the first time genetic parameters and G x E for pearl quality traits when multiple families were again reared at Bali and Lombok. Here significant differences in the size and value of pearls produced at the two locations were observed, with pearls produced at Lombok generally bigger and more valuable than their Bali counterparts. Comparisons of pearls produced by the various families jointly reared at these two sites also indicated adverse genetic correlations for size (r(g) = -0.22), colour (r(g) = 0.28) and weight (r(g) = 0.38), and less so shape (r(g) = 0.56) and lustre (r(g) = 0.59); thus the occurrence of genotype by environment modifications for these pearl quality traits needs to be factored into improvement programs. Heritability analyses based on the donor-oyster additive genetic contribution showed that all pearl traits except that of shape exhibited low to moderate heritabilities (size h² = 0.13, lustre h² = 0.14, weight h² = 0.15, colour h² = 0.15 and complexion h² = 0.25). As a consequence these traits could be improved through the practice of selection. The findings of this thesis have substantially advanced our knowledge of the respective role genetics and the environment play in the realization of commercially important traits in the pearl oyster P. maxima. Pearl quality and oyster growth traits have been shown to have a heritable basis, thereby making them amiable to improvement through selection approaches. Results also have shown that when designing future breeding programs considerations of large-scale site induced environmental effects and associated genotype by environment modifications will need to be factored. Through implementation of the information gathered in this thesis the P. maxima pearling industry now has a sound basis for the future design of efficient selection programs aimed at improving the productivity and profitability of their industry