Luận án tiến sỹ về di truyền và sinh sản hàu (PhD thesis 2016)

Luận án nghiên cứu về di truyền hàu làm cơ sở cho thiết lập chương trình chọn giống hàu như: định danh chính xác tên loài, xác định hệ số di truyền liên quan đến quần đàn chọn giống, và xác định các peptide hóc môn có vai trò trong phát triển tuyến sinh dục của hàu

Literature Review

Introduction

Genetic programs are an essential component of aquaculture operations being a management tool not only to prevent inbreeding but also to allow selection for improved traits and increase profitability A key requirement for the operation of genetic programs is the ability to reproduce select animals in captivity, that is, to have full control of the reproductive process This study will focus on genetics and reproduction of two important aquaculture species: Pacific oyster, Crassostrea gigas / Portuguese oyster, C angulata (VNO) and Sydney rock oyster, Saccostrea glomerata with its aim to develop genetic, molecular and physiological tools required for an optimal operation of genetic programs and apply those to the VNO and S glomerata Scientifically, it will provide knowledge about how long term selection affects genetic diversity, about the endocrine system controlling reproduction in oysters and how this information can be used to permit reliable reproduction and set up sustainable genetic improvement programs To achieve the above objectives, four experiments were implemented including: 1) Analysis of selected S glomerata for genetic diversity; 2) Identification of oyster populations cultured in Vietnam and analysis of their genetic diversity; 3) Application of the genetic and physiological tools to design a trial breeding program for the VNO in Vietnam; 4) Identification of key genes and peptide hormones associated with reproductive performance of S glomerata

Oyster production

In 2010, world oyster production reached approximately 5 million tonnes (FAO, 2014) In Australia, it is one of the oldest aquaculture species with a history of 120 years (Nell, 2005), mainly present in South Australia, New South Wales, Tasmania and to a lesser extent in Queensland and Western Australia (Maguire and Nell, 2005) with two main species: Sydney rock oyster, S glomerata and Pacific Oysters, C gigas

C gigas, native to Japan, was introduced to mainland Australia in the late 1940s and to

Tasmania in the early 1950s; it then quickly established wild populations in Australia (English,

2000, Ward et al., 2000) C gigas was introduced to France in the early 1970s because of the mass mortality of C angulata (Boudry, 2008) Curently, it is now being cultured in 27 countries and is the most highly produced mollusc species in the world (Boudry, 2008) as C gigas has relatively high disease resistance and fast growth compared to other oyster species The average time for S glomerata to grow to a marketable size (40-60g) is about 3.5 years , whereas it only takes about 17-18 months for C gigas to reach market size in Tasmania (Maguire and Nell, 2005) and as short as 12 months in NSW

In Vietnam, before 2008, mollusc production was dominated by the aquaculture of hard clams (Meretrix meretrix and M lyrata) with seed collected from the wild; other mollusc species accounted for only a small proportion of the total Vietnamese production (RIA1 annual report,

2012) In 2012, the total mollusc production was appropximately 190,000 t/annum (FIS/2010/100, 2013) The so-called “Pacific Oyster” (VNO) was introduced into Vietnam for both hatchery and grow-out production in 2008 via the support of an ACIAR project (Building bivalve mollusc hatchery production capacity project in Vietnam and Australia, FIS/2005/114) The growth of VNO production in Vietnam was very impressive, rising from no production in 2006-

07 to 7000 t by 2010-11 , worth A$7.0-9.8 million annually (FIS/2010/100, 2013) Currently, VNO production consists of approximately 200 small farms, operating 2,200 rafts in northern Vietnam (FIS/2010/100, 2013) One of the major constraints in VNO production in Vietnam is lack of reliable seed supply, poor quality of spat and quality assurance to meet the growing demands of farmers and producers (FIS/2010/100, 2013)

Reproduction

One of the key requirements for the operation of genetic programs is the ability to reproduce selected animals in captivity, that is, to have full control of the reproductive process However, the molecular and biochemical cues associated with oyster reproduction are poorly understood Therefore, identification of genes and peptide hormones that regulate oyster reproduction will help establish knowledge for a better understanding of this process Through transcriptome analysis, I aimed to identify genes and gene products that can trigger reproduction and investigate how this knowledge could be applied to optimize the design for a genetic breeding programs

Oysters obtain food by filtering and extracting algae, bacteria and nutrients from the surrounding water Most oyster species, including S glomerata and C gigas/VNO, change sex during their life The first spawning is usually as a male and subsequent spawning as a female while some remain as hermaphrodites In adult S glomerata populations, the percentage of males and females can differ greatly, for example 33% and 67% (Dove and O'Connor, 2012) During spawning, adult S glomerata females can spawn up to 20 million eggs (O'Connor et al., 2008b), while a large C gigas may release up to 100 million eggs per spawn and males release hundreds of millions of sperm into the water when the tides and currents are optimal for the widest distribution Fertilization takes place in the water column and embryo development continues for up to 3-4 weeks as the larval stages of the oyster swim and grow, ultimately settling on a suitable hard surface, then never leaving their settled position

The development of oysters has been described in terms of five developmental phases, where phase I is the ripening period, phase II is fully ripe, phase III is post spawning, phase IV is regression with the presence of phagocytes, and phase V is a regressive phase where gonial cells are indeterminate (Dinamani, 1974) The stages of development of bivalves are shown in Figure 1.1 The development of embryos to D-veliger larvae of S glomerata is saffected by temperature, salinity and the interaction of these factors Salinity of 35 ppt and 26°C are optimal for embryonic development whereas D-veliger larvae grow best at 28°C and at maximum salinity level of 34 ppt The highest growth rate of spat was observed at a salinity of 35 ppt and a temperature of 30°C (Dove and O'Connor, 2007)

The development of embryos and larvae for both C glomerata, C gigas are well-known and recorded in manuals for hatchery production (Dove and O'Connor, 2012) However, the information on regulation of oyster reproduction or sex changes is very limited In molluscs, neuropeptides and peptide hormones are known to have multiple functions from physiology, behaviour and reproduction Therefore, the identification and investigation of functions of neuropeptides involved in oyster reproductive regulation will greatly support reliable hatchery production

Figure 1.1 Stages of development in bivalves (Helm et al., 2004)

In bivalve molluscs, neurohormones are generated in ganglia of the central nervous system that consists a large number of neurosecretory cells There are two main parts: the cerebral ganglia and the visceral ganglia (Figure 1.2) where the visceral ganglia are much larger than the cerebral ganglia Several vertebrate-like hormone immunoreactivities have been found in the mussel ganglia: insulin-like, α-MSH-like, CCK-like, somatostatin-like, FMRFa-like, substance P-like, and neuropeptide F-like However, their involvement in the reproductive process of molluscs is still unknown (Pazos and Mathieu, 1999) The role of the various other peptides reported in the endocrine system for the regulation of gonad development, spawning behaviour and sex changes is unclear and needs to be examined

Figure 1.2 The nervous system of Crassostrea viewed from the right (Fox, 2007)

Genetic variation and genetic programs

One of the problems in oyster industries is degraded stock due to inbreeding (Knibb et al., 2014b) For sustainable culture, a selective breeding program which delivers genetic

24 improvement for specific traits, yet prevents inbreeding, is indispensable Maintaining a wide range of genotypes could give a hatchery population more flexibility of response to constantly changing environments (Boudry, 2008, Taris et al., 2006) Genetic diversity is the main resource for a stock improvement program, however it may be eroded due to domestication selection, or poor husbandry practices in hatcheries where often a very limited number of broodstock individuals are kept resulting in high inbreeding and deterioration of quality seed stock (Boudry,

2008, Nguyen, 2009, Taris et al., 2006) Loss of genetic variation and inbreeding may then lead to poor future genetic gains in a breeding program (In et al., 2016a) Inbreeding depression may reduce fitness (ability to survive and reproduce) and production performance of the animals Therefore, maintenance of additive genetic variance is important as it provides bbetter selection repsonse over generations of selective breeding

Growth rate and disease resistance can be improved through a genetic program as heritable variation exists for these traits (Boudry et al., 1997) Mass selection and family based selection are two kind of approaches in selective breeding Mass selection is suitable only when a single trait is chosen for improvement, such as fast growth; meanwhile other traits especially with low or moderate heritability such as shell shape may be better improved through family selection (Ward et al., 2000) To date, breeding programs for S glomerata and C gigas have been set up in many countries including Australia, New Zealand, Korean, China and elsewhere in the world to improve growth rate and disease resistance (Appleyard and Ward, 2006, Hwang et al.,

2013, Kaspar et al., 2013, Li et al., 2013, O'Connor and Dove, 2006) The main traits of interest for a genetic program have been fast growth Selection for fast growth has been successful for several oyster species including Crassostrea virginica, Ostrea edulis and Ostrea chilensis (Dove and O'Connor, 2009, Nell et al., 1999, O'Connor and Dove, 2006)

Specifically for C gigas, selection for fast growth was initated in Tasmania, Australia in

1996 (Appleyard and Ward, 2006) and subsequently selective breeding developed fast growth and disease resistant stocks However, a high level of inbreeding due to mating of closely related individuals could result in depression of performance characteristics (Evans et al., 2004b) This resulted in a decline in response on first or second generation (Evans et al., 2004b, Toro and Newkirk, 1990) In order to balance maximum selection intensity and the need to retain an adequate number of broodstocks, inbreeding accumulation of no more than 1% is considered acceptable as it may not adversely affect performance gains too much (Falconer and Mackey,

1996, Frankel and Soule, 1981) Specifically for S glomerata, the third generation stock from selection for high growth was 74% heavier (50g greater) than controls which saves up to 11 months (28 for selected vs 38 months for non-selected oysters) of the time otherwise taken to reach the market size (Nell, 2005, Nell, 2006)

However other traits such as disease resistance, shape and uniformity are increasingly becoming important (Appleyard and Ward, 2006, Kube et al., 2013) C gigas mortality syndrome (POMS) occurred and caused mass mortality in Europe, New Zealand, Korea and Australia during summer times raising the need to develop POMS resistant stocks for the C gigas industry (Hwang et al., 2013, Kube et al., 2013) In Australia, the resistance to POMS has been added in the breeding objective for C gigas in order to produce resistant stock for growers (Kube et al., 2013)

In S glomerata, Winter mortality syndrome (Bonamia roughleyi) and QX disease (Marteilia sydneyi) has caused more than 40% reduction in S glomerata production in Australia (Heasman et al., 2000, Nell, 2003) QX may kill 80% or more of the infected S glomerata population (Nell and Perkins, 2006, Simonian et al., 2009) However, the mortality of both diseases was reduced by selective breeding conducted by the NSW DPI (Simonian et al., 2009) After two generations of selection, the new strains resistant to QX (Marteilia sydneyi) and winter mortality diseases had 29% less mortality from QX disease at Lime Kiln Bar, Georges River (Nell, 2003) Selective breeding was also developed for fast growth and MSX disease resistance in lines of American Oyster (Crassostrea virginica) in the United States in order to avoid high mortality rates from MSX disease (Standish et al., 1993)

Understanding the level of genetic variation for traits of commercial importance and their genetic correlations is important in selective breeding programs as it shapes how to do selection (what selection indices to use, what weights to put on particular traits) This involves the estimation of genetic parameters such as heritabilities (h 2 ) and genetic correlations High heritabilities for a trait of interest show that a large proportion of the phenotypic variation is due to genetics (Falconer, 1981, Newkirk et al., 1977) Heritability values of 0.20 or larger indicate that genetic improvement can easily be achieved through a selective breeding program (Newkirk et al., 1977) Correlation implies that the alteration of one trait will cause correlated changes in other traits (Falconer, 1981) The changes can be either positive or negative The magnitude of changes will be affected by degree of correlation between the traits involved (Toro and Newkirk,

In summary, for both C gigas and S glomerata, there are a limited number of genetic tools available to calibrate and manage diveristy in hatchery stocks There is also limited knowledge of the molecular and biochemical cues associated with reproduction The genetic diversity studies will determine if current breeding practices for C gigas and S glomerata are sustainable in the long term The discovery of bioactive compounds associated with reproduction will support more reliable reproduction, which in turn will support the genetic breeding programs.

DNA barcoding for resolving close-related oysters

In Vietnam, an industry based on a newly introduced oyster, called “Pacific oyster” (VNO) started in 2008, however, it has not been clear whether the VNO is C gigas This is because they were imported from a population in Taiwan which is known as a pure stock of C angulata

(Boudry et al., 1997) In addition, Vietnamese farmers also imported oyster spat from South of China which are assumed C angulata but they may consist of C gigas or hybrid stocks Unfortunately, C angulata and C gigas are the genetically very close taxa which are difficult to identify using morphology and physiology criteria (Huvet et al., 2000b) In the past, they were classified as two different species by Thunberg in 1793 and Lanmarck in 1819 as they apparently were distributed in two separate areas: C angulata in Europe and C gigas in Asia (Lapegue et al., 2004) However, the two taxa were then considered as a single species due to high genetic similarity between the taxa based on morphologic e.g Boudry et al (1998) comparision, experimental hybridization (Huvet et al., 2001, Huvet et al., 2002) and allozyme studies In the past, mitochondrial DNA (mtDNA) studies were used for clarification of taxonomic status and analysis of stock structure in some commercially important oysters (Klinbunga et al., 2003) Boudry et al (1998) and Huvet et al (2000b) suggested mtDNA show clear differences between

C gigas and C angulata and thus mtDNA sequencing may be an effective tool for identification of oyster populations in Vietnam

Specifically for differentiation of C.gigas and C.angulata, allozyme and molecular DNA markers such as microsatellites and mitochondrial DNA (mtDNA) have been used mtDNA is sometimes considered as preferable to nuclear DNA for phylogenetic studies due to its uniparental inheritance, high evolutionary rate, lack of introns, large copy numbers in every cell and limited recombination (Radulovici et al., 2010)

Previous studies using protein allozyme markers supported the hypothesis that the two species should be classified as the same species (Boudry et al., 1998) Similarly, nuclear markers did not help clarify the two taxa Huvet et al (2000a) and Reece et al (2008) However, mtDNA has revealed possible genetic differences between them and proposed mtDNA can be used to distinguish these taxa (Boudry et al., 1998, Huvet et al., 2000b) The differentiation between the taxa tended to increase considering mtDNA sequences rather than nuclear sequences (Huvet et al., 2000a) This can be explained by the fact that mtDNA is often more polymorphic than nuclear DNA since the former has a relatively faster mutation rate, which results in greater variation between species Indeed recently, mtDNA sequences were used to consider relationships among six Asian oysters including C gigas and C angulata (Wu and Yu, 2009)

Among genes of the mtDNA chromosome, the mitochondrial cytochrome C oxidase subunit I (COX1) gene is one of the most commonly used markers in molecular studies and barcoding as it provides strong phylogenetic differentiation(Hebert et al., 2003) It was widely used as a DNA barcode for attempted phylogenetic separation of C angulata and C gigas Other mtDNA genes: rrnL and MNR, 16S rDNA, 12S rDNA (David and Savini, 2011, Lam and Morton,

2006, Masaoka and Kobayashi, 2005, Stepien et al., 2001) and nuclear genes: ITS1, ITS2 18S and 28S rDNA (David and Savini, 2011, Larsen et al., 2005) were also used to analyze the phylogeny of bivalves but they are less diverse than COX1 (Boudry et al., 2003, Radulovici et al., 2010, Stepien et al., 2001) Sequence comparison of the Internal transcribed spacer (ITS) region is another tool for taxonomy and molecular phylogeny Although ITS has high degree of variation among some species of Ostreidae, it could not separate the closely related species such as C gigas and C angulata (Wang and Guo, 2008)

Boudry et al (1998) used mtDNA-RFLP for analysis of COX1 of C gigas and C angulata with TaqI, MseI, Sau3AI, HhaI to reveal 6 mitotypes composed of A (ccab), B(cdab), C(dcad),

D(dcab), E(dcbd), and J(acab) showing 76% of C gigas contain C mitotypes, but 88% of C angulata hold A mitotypes COX1 was also considered an effective tool to identify the European flat oyster Ostrea edulis, the native O angasi in Oyster Harbour, Western Australia (Morton et al., 2003), C iredalei and Saccostrea cucullata in Thailand (Klinbunga et al., 2003), Malaysian

Crassostrea oyster species including Crassostrea iredalei, Crassostrea belcheri and Crassostrea madrasensis (Mustaffa et al., 2010) and to discover new oyster species, such as Crassostrea hongkongensis (Lam and Morton, 2003)

Therefore, the principal aim of this study/thesis was to establish foundational knowledge to develop an optimal genetic program for oysters in Australia and Vietnam The study involved four major experiments: 1) Identify genes and peptide hormones that regulate oyster reproduction, 2) Develop the molecular tools and genetic markers to analyse genetic diversity for selected S glomerata and hatchery VNO stocks, 3) Investigate genetic integrity of VNO to understand whether they are C gigas or C angulata and whether they have adequate genetic variation to start a breeding program The fourth experiment was to evaluate strain genetic effects and heterotic expression of selected VNO in various culture systems and different rearing environments

Specifically, the thesis includes the six chapters that have been prepared in the format of four papers as the followings:

2) Chapter 2 : Vu Van In, Nikoleta Ntalamagka, Wayne O’Connor, Tianfang Wang, Dan Powell, Scott

F Cummins and Abigail Elizur Reproductive neuropeptides that stimulate spawning in the Sydney rock oyster, Saccostrea glomerata Peptide

3) Chapter 3 : Vu Van In, Wayne O’Connor, Michael Dove, Wayne Knibb 2016 Can genetic diversity be maintained across multiple mass selection lines of Sydney rock oyster, Saccostrea glomerata despite loss within each? Aquaculture, 454, 210-216

4) Chapter 4 : Vu Van In, Phan Thi Van, Vu Van Sang and Wayne Knibb Is the Vietnam aquaculture pacific oyster Crassostrea gigas ? And can aquaculture practice remain gentic variation among?

A case study on taxonomy and genetic diversity for a start of breeding program To be submitted to Aquaculture

5) Chapter 5: Vu Van In, Phan Thi Van, Nguyen Thi Thu Hien, Wayne Knibb and Nguyen Hong

Nguyen Are strain genetic effect and heterosis expression altered with culture system and rearing environment in Portuguese oyster, Crassostrea angulata ? Aquaculture Research, 2016 1-12

References

APPLEYARD, S A & WARD, R D 2006 Genetic diversity and effective population size in mass selection lines of Pacific oyster (Crassostrea gigas) Aquaculture, 254, 148-159

BOUDRY, P 2008 Review on breeding and reproduction of Europhean aquaculture species Aqua

BOUDRY, P., BARRE, M & GERARD, A Genetic improvement and selection in shellfish: A review based on oyster research and production 1997

BOUDRY, P., HEURTEBISE, S., COLLET, B., CORNETTE, F & GÉRARD, A 1998 Differentiation between populations of the Portuguese Oyster, Crassostrea angulata and the Pacific Oyster,

Crassostrea gigas, revealed by mtDNA RFLP analysis Journal of Experimental Marine Biology and Ecology, 226 , 279-291

BOUDRY, P., HEURTEBISE, S & LAPÈGUE, S 2003 Mitochondrial and nuclear DNA sequence variation of presumed Crassostrea gigas and Crassostrea angulata specimens: a new oyster species in Hong Kong? Aquaculture, 228 , 15-25

DAVID, D.-C & SAVINI, D 2011 Molecular approaches to bivalve population studies: a review

DINAMANI, P 1974 Reproductive cycle and gonadial changes in the New Zealand Rock Oyster,

Crassostrea glomerata New Zealand Journal of Marine and Freshwater Research, 8, 39-65

DOVE, M C & O'CONNOR, W A 2007 Salinity and temperature tolerance of Sydney Rock Oysters

Saccostrea glomerata during early ontogeny Journal of Shellfish Research, 26 , 939-947 DOVE, M C & O'CONNOR, W A 2009 Commercial assessment of growth and mortality of fifth generation Sydney Rock Oysters, Saccostrea glomerata (Gould, 1850) selectively bred for faster growth Aquaculture Research, 40 , 1439-1450

DOVE, M C & O'CONNOR, W A 2012 Reproductive cycle of Sydney Rock Oysters, Saccostrea glomerata (Gould 1850) selectively bred for faster growth Aquaculture, 324–325, 218-225

ENGLISH, L J., MAGUIRE, G B., WARD, R D 2000 Genetic variation of wild and hatchery populations of the Pacific Oyster, Crassostrea gigas (Thunberg), in Australia Aquaculture, 187, 283-298 EVANS, F., MATSON, S., BRAKE, J & LANGDON, C 2004 The effects of inbreeding on performance traits of adult Pacific oysters, Crassostrea gigas Aquaculture, 230 , 89-98

FALCONER, D S 1981 Introduction to quantitative genetics, 2 nd edition Logman Group Ltd New

FALCONER, D S & MACKEY, T F C 1996 Introduction to Quantitative Genetics Longman, Essex,

FAO 2014 Food and Agriculture Organization (FAO) of the United Nations 2014 Yearbook of

Fisheries Statistics extracted with FishStatJ (Copyright 2013) Fisheries database: Aquaculture production quantities 1950-2012; aquaculture production values 1984-202; capture production 1960-2012; Commodities Production and Trade 1976-2011 http://www.fao.org/fishery/statistics/en Accessed October 2014

FIS/2010/100 2013 Enhancing bivalve production in northern Vietnam & Australia Aciar project

FOX, R 2007 Invertebrate Anatomy Online, Crassostrea virginica, American Oyster The American

Oyster, Crassostrea virginica [Online] [Accessed 2007]

FRANKEL, O H & SOULE, M E 1981 Conservation and Evolution Cambridge University Press,

HEASMAN, M P., GOARD, L., DIEMAR, J & R.B, C 2000 Improved early survivals of mollusks: Sydney

Rock Oyster, NSW Fishereis Report, NSW, Australia: Port Stephens Fishereis Center, 2003 HEBERT, P D., CYWINSKA, A & BALL, S L 2003 Biological identifications through DNA barcodes

Proceedings of the Royal Society of London Series B: Biological Sciences, 270, 313-321

HELM, M M., BOURNE, N & LOVATELLI, A 2004 Hatchery Culture of Bivalves: A Practical Manual

FAO Fisheries Technical Paper No 471 FAO, Rome 177 pp

HUVET, A., BALABAUD, K., BIERNE, N & BOUDRY, P 2001 Microsatellite Analysis of 6-Hour-Old

Embryos Reveals No Preferential Intraspecific Fertilization Between Cupped Oysters

Crassostrea gigas, Crassostrea angulata Marine Biotechnology, 3, 448-453

HUVET, A., BOUDRY, P., OHRESSER, M., DELSERT, C & BONHOMME, F 2000a Variable microsatellites in the Pacific Oyster Crassostrea gigas and other cupped oyster species

HUVET, A., GÉRARD, A., LEDU, C., PHÉLIPOT, P., HEURTEBISE, S & BOUDRY, P 2002 Is fertility of hybrids enough to conclude that the two oysters Crassostrea gigas and Crassostrea angulata are the same species? Aquatic Living Resources, 15, 45-52

HUVET, A., LAPEGUE, S., MAGOULAS, A & BOUDRY, P 2000b Mitochondrial and nuclear DNA phylogeography of Crassostrea angulata, the Portuguese Oyster endangered in Europe

HWANG, J Y., PARK, J J., ARZUL, I & PARK, M A 2013 Ostreid herpesvirus infection in farmed Pacific

Oyster larvae, Crassostrea gigas in Korea 5th International Oyster Symposium (IOS5), World

Oyster Society Ho Chi Minh city, December 2013

IN, V.-V., O'CONNOR, W., DOVE, M & KNIBB, W 2016 Can genetic diversity be maintained across multiple mass selection lines of Sydney rock oyster, Saccostrea glomerata despite loss within each? Aquaculture, 454 , 210-216

KASPAR, H F., JANKE, A R., CAMARA, M D., KING, N., YEN, S & RAGG, N L C 2013 5 th International

Oyster Symposium (IOS5), World Oyster Society Ho Chi Minh city, December 2013

KLINBUNGA, S., KHAMNAMTONG, N., TASSANAKAJON, A., PUANGLARP, N., JARAYABHAND, P &

YOOSUKH, W 2003 Molecular Genetic Identification Tools for Three Commercially Cultured Oysters (Crassostrea belcheri, Crassostrea iredalei, and Saccostrea cucullata) in Thailand

KNIBB, W., WHATMORE, P., LAMONT, R., QUINN, J., POWELL, D., ELIZUR, A., ANDERSON, T.,

REMILTON, C & NGUYEN, N H 2014 Can genetic diversity be maintained in long term mass selected populations without pedigree information? A case study using banana shrimp

KUBE, P., HICK, P., CUNNINGHAM, M., KIRKLAND, P., ELLIOTT, N., O’CONNOR, W & DOVE, M C 2013

Current progressin genetic selection for resistance to acific Oyster mortality syndrome caused by OsHV-1 microvariant (μ-var) in Australia 5 th International Oyster Symposium (IOS5), World Oyster Society Ho Chi Minh city, December 2013

LAM, K & MORTON, B 2003 Mitochondrial DNA and morphological identification of a new species of Crassostrea (Bivalvia: Ostreidae) cultured for centuries in the Pearl River Delta, Hong Kong, China Aquaculture, 228 , 1-13

LAM, K & MORTON, B 2006 Morphological and mitochondrial-DNA analysis of the Indo-West Pacific rock oysters (Ostreidae: Saccostrea species) Journal of molluscan studies, 72, 235-245 LAPEGUE, S., BATISTA, F., HEURTEBISE, S., YU, Z & BOUDRY, P 2004 Evidence for the presence of the Portuguese oyster, Crassostrea angulata in northern China Journal of Shellfish Research,

LARSEN, J B., FRISCHER, M E., RASMUSSEN, L J & HANSEN, B W 2005 Single-step nested multiplex

PCR to differentiate between various bivalve larvae Marine Biology, 146 , 1119-1129

LI, Q., KONG, N., YU, H & KONG, L 2013 Heritability estimates for growth-related traits in the Pacific

Oyster Crassostrea gigas using a molecular pedigree 5 th International Oyster Symposium (IOS5), World Oyster Society Ho Chi Minh city, December 2013

MAGUIRE, G B & NELL, J A 2005 History, status and future of oyster culture in Australia The 1 st

International Oyster Symposium Proceedings, Oyster Research Institute News No.19

MASAOKA, T & KOBAYASHI, T 2005 Species identification of Pinctada imbricata using intergenic spacer of nuclear ribosomal RNA genes and mitochondrial 16S ribosomal RNA gene regions

MORTON, B., LAM, K & SLACK-SMITH, S 2003 First report of the European flat oyster Ostrea edulis, identified genetically, from Oyster Harbour, Albany, south-western Western Australia

MUSTAFFA, S., L., A M., ARIFFIN, A H., M.N., D & NOR, S A M 2010 Cytochrome oxidase I (COI) gene highlights taxonomic ambiguities of Malaysian Crassostrea (Sacco, 1897) oyster species The

7th IMT-GT UNINET and the 3rd Joint International PSU-UNS Conferences, Proceedings, 7-8 October 2010, Prince of Songkla University, Hat Yai, Songkhla, Thailand / Prince of Songkla University

NELL, J 2005 Primefacts 3: Farming the Sydney rock oyster NSW Department of Primary Industries,

NELL, J A 2003 Selective breeding for disease resistance and fast growth in Sydney rock oysters , NSW

NELL, J A 2006 Manual for mass selection of Sydney rock oysters for fast growth and disease resistance Fisheries Research Report series 13 NSW: NSW Department of Primary Industries NELL, J A & PERKINS, B 2006 Evaluation of the progeny of third ‐ generation Sydney Rock Oyster,

Saccostrea glomerata (Gould, 1850) breeding lines for resistance to QX disease Marteilia sydneyi and winter mortality Bonamia roughleyi Aquaculture Research, 37, 693-700

NELL, J A., SMITH, I R & SHERIDAN, A 1999 Third generation evaluation of Sydney rock oyster

Saccostrea commercialis (Iredale and Roughley) breeding lines Aquaculture, 170 , 195-203 NEWKIRK, G F., HALEY, L E., WAUGH, D L & DOYLE, R 1977 Genetics of larvae and spat growth rate in the oyster Crassostrea virginica Marine Biology 41 , 49-52

NGUYEN, D 2009 Assessing genetic diversity in cultured aquatic species: the Sydney Rock Oyster,

Saccostrea glomerata stock improvement program as a model Masters by research thesis,

O'CONNOR, W & DOVE, M 2006 Reproductive conditioning of oysters selected for fast growth

Australasian Aquaculture 2006 27-30 August 2006, Adelaide, Australia

O'CONNOR, W A., DOVE, M C., FINN, B & O’CONNOR, S J 2008 Manual for hatchery production of sydney rock oysters, Saccostrea glomerata IDP NSW NSW Department of Primary Industries Port Stephens Fisheries Center Taylors Beach NSW 2316, Australia http://www.dpi.nsw.gov.au/research/areas/aquaculture/outputs/2008/oconnor3

PAZOS, A & MATHIEU, M 1999 Effects of five natural gonadotropin-releasing hormones on cell suspensions of marine bivalve gonad: stimulation of gonial DNA synthesis General and

RADULOVICI, A E., ARCHAMBAULT, P & DUFRESNE, F 2010 DNA Barcodes for Marine Biodiversity:

REECE, K S., CORDES, J F., STUBBS, J B., HUDSON, K L & FRANCIS, E A 2008 Molecular phylogenies help resolve taxonomic confusion with Asian Crassostrea oyster species Marine Biology, 153, 709-721

RIA1 ANNUAL REPORT 2012 Bivalve production in Vietnam Annual report

SIMONIAN, M., NAIR, S V., NELL, J A & RAFTOS, D A 2009 Proteomic clues to the identification of

QX disease-resistance biomarkers in selectively bred Sydney Rock Oyster, Saccostrea glomerata Journal of Proteomics, 73 , 209-217

STANDISH, K., ALLEN, J., PATRICK, M G & JOHN, W E 1993 Genetic improvement of the Eastern oyster for growth and disease resistance in the Northeast NRAC Fact Sheet N.210 - 1993 University of Massachusetts Dartmouth North Dartmouth Massachusetts 20747

STEPIEN, C., MORTON, B., DABROWSKA, K., GUARNERA, R., RADJA, T & RADJA, B 2001 Genetic diversity and evolutionary relationships of the troglodytic ‘living fossil’Congeria kusceri (Bivalvia: Dreissenidae ) Molecular Ecology, 10 , 1873-1879

TARIS, N., ERNANDE, B., MCCOMBIE, H & BOUDRY, P 2006 Phenotypic and genetic consequences of size selection at the larval stage in the Pacific oyster, Crassostrea gigas Journal of Experimental Marine Biology and Ecology, 333 , 147 - 158

TORO, J E & NEWKIRK, G F 1990 Divergent selection for growth rate in the European oyster Ostrea edulis : response to selection and estimation of genetic parameters Marine Ecology Progress series, 62, 219-227

WANG, Y & GUO, X 2008 ITS length polymorphism in oysters and its use in species identification

WARD, R D., ENGLISH, L J., MCGOLDRICK, D J., MAGUIRE, G B., NELL, J A & THOMPSON, P A 2000

Genetic improvement of the Pacific oyster Crassostrea gigas (Thunberg) in Australia

WU, X & YU, Z 2009 Comparative mitogenomic analyses and phylogenetic investigation in oysters from Crassostrea , Saccostrea and Ostrea The 3rd International Oyster Symposium (IOS3) Nov

Reproductive neuropeptides that stimulate spawning in the Sydney rock oyster

Background

The Sydney rock oyster, Saccostrea glomerata, is one of the most ecologically and commercially important species of the oyster family (Ostreidae) in Australian waters In the wild, it dominates sheltered shorelines of intertidal and immediate subtidal regions along the Eastern Australian coast It also forms the basis of an extensive oyster industry in South-East Queensland and New South Wales S glomerata production is one of the oldest aquaculture industries in Australia and its current production has reached 7,793,390 dozen of oysters (at farm gate), valued at around $34.7 million for 2014/2015 (Trenaman et al., 2015)

Critical to the production and marketing of S glomerata is their physical and reproductive condition S glomerata is a protandric species, where the gonad condition cycles are broadly understood, and to some extent can be manipulated through gross environmental changes (O'Connor et al., 2008a, O'Connor et al., 2008b) However, our understanding of the molecular and biochemical processes underpinning changes in gonadal condition, as well as our capacity to monitor these changes, is limited Thus, our ability to manipulate maturation and spawning in S glomerata is also limited The identification of neuropeptides that may regulate S glomerata reproduction provides an important research area which should contribute to our understanding of S glomerata biology as well as help to facilitate hatchery production (Dove and O'Connor,

Neuropeptides are produced and released by neurons through a regulated secretory pathway (Burbach, 2011) They represent a highly diverse and multifunctional group of signalling molecules that include hormones, neurotransmitters and neuromodulators (Conzelmann et al.,

2013, Stewart et al., 2014) Their roles in the molluscan physiology, behaviour and reproduction are well established (Fricker, 2012, Morishita et al., 2010), and includes APGWamide, gonadotropin releasing hormone (GnRH), and egg-laying hormone (ELH), which have each been investigated through in vitro or in vivo studies These neuropeptides are generated from precursor sequences (Cummins et al., 2011, Morishita et al., 2010, Nuurai et al., 2010)

In the oysters, it has been demonstrated that the tetrapeptide APGWa plays a role in the reproduction of Crassostrea gigas female oysters, where it can induce in vitro adductor muscle contraction followed by oocyte release (Bernay et al., 2006) GnRH is a well-known reproductive regulator in vertebrates, but has also been found in the CNS of bivalve molluscs (Bigot et al.,

2012, Nakamura et al., 2007 GnRH-like peptides have been identified in the visceral, cerebral and pedal ganglia of scallops Patinopecten yessoensis, the oyster: C gigas and P fucata (Treen,

2012, Pazos and Mathieu, 1999, Stewart et al., 2014) In support of their role in reproduction, there exists a highly expressed orthologue of the GnRH receptor in mature gonads of C gigas (Morishita et al., 2010) Also, in vitro trials in C gigas and the mussel Mytilus edulis show that GnRH stimulates proliferation of gonadal cells (Pazos and Mathieu, 1999) Finally, the ELH is a well-known spawning inducer in the aquatic snails Aplysia and Lymnaea (Conn and Kaczmarek,

1989, Strumwasser et al., 1987) and its gene sequence has been identified in C gigas and P fucata (Stewart et al., 2014) Numerous other molluscan neuropeptides exist, although their role in reproduction is less well known

The main objective of the present study was to identify neuropeptide genes that may be important in regulating S glomerata reproduction We focused on the neural and gonad tissues through transcriptome and peptidomic analysis, and then performed in vivo maturation and spawning bioassays to elucidate potential reproduction-associated roles.

Materials and methods

Animals for tissue collection for RNA

Wild live adult S glomerata were obtained from Port Stephens Fisheries Research Institute, New South Wales (PSFI) The stage of gonadal development of each individual was determined (stages I-V) as described by Dinamani (1974) Gonad and visceral ganglia tissues were isolated from each individual within three out of five gonadal development stages: 1) stage I - Ripening; 2) stage II - Fully ripe and 3) Stage III - Post spawning (20-25 oysters/each stage, Np) in July, 2012 Tissues from males and females were kept separately for gonad and ganglia at -

80 o C until used for total RNA and peptide extraction

Live adult S glomerata from wild and hatchery lines were obtained from either local retail outlets on the Sunshine coast, QLD (for bioassays carried out at University of the Sunshine Coast, USC) or from PSFI (for those carried out at PSFI) Animals were acclimatized in culture tanks for at least 24 h, fed with algae before used for the experiments

The overall procedure applied in this study is shown in Figure 2.1

Figure 2.1 Workflow of the experiments to identify neuropeptides of S glomerata using transcriptomics and peptidomics Position of the visceral ganglia and gonad is shown in schematic [modified from (Paul, 1964)]

Detecting peptides in each fraction List of posible peptide precursors

RNA extraction and transcriptome sequencing

Total RNA was isolated using TRIsure TM Reagent (Bioline USA Inc.) following the manufacturer’s specifications The quality and concentration of the total RNA were checked by gel electrophoresis and spectrophotometry (Nanodrop 2000, Thermo Scientific, USA) Total RNA of each sex was pooled separately from all developmental stages (stages I-V) Twenty micrograms of total RNA of each tissue were freeze dried and sent to BGI for de novo transcriptome sequencing BGI conducted RNA-Seq with 2 lanes of 100 paired-end sequencing on the illumine HiSeq2000, assembly and functional annotation De novo assemblies were performed by SOAPdenovo software using trimmed reads from the Illumina sequencing The assembler was run with the parameter sets as following: seqType, fq; minimum kmer coverage

= 4; minimum contig length of 100 bp; group pair distance = 250

To identify target sequences, gender-specific transcriptomes for the gonadal and visceral ganglia of S glomerata were imported into the CLC Main Workbench (v7.0.2; CLC-bio, Denmark) Previously identified molluscan neuropeptide sequences were then queried (tBLASTn) against the transcriptomes To complement this, Open Reading Frames (ORF) were retrieved from the S glomerata databases and screened for signal sequences using SignalP 4.0 Server

(http://www.cbs.dtu.dk/services/SignalP-4.0/) The presence of recurrent KK; KR; RK cleavage sites was identified using NeuroPred (Web-based software on http://neuroproteomics.scs.illinois.edu/neuropred.html) Multiple sequence alignments were done by MEGA software version 6.06 (Tamura et al., 2013) Derived and amino acid sequences were aligned, guided by chain cleavage sites and conserved cysteines (Brunak et al., 1991) Domain graph 2.0 and Miktex-2.9 were used to build the peptide schematics and sequence alignments, respectively Data of other species used for alignment of amino acid sequences and schematic diagrams was obtained from the supplementary list of neuropeptides provided by Stewart et al (2014) Web-based Clustal Omega (http://www.ebi.ac.uk/Tools/msa/clustalo/) was used to estimate percentage of identity between S glomerata peptide amino acid sequences and other mollusc species

Reverse phase-high performance liquid chromatography (RP-HPLC)

The collected ganglia tissues were homogenized on ice in a solution of 0.1% Trifluro-acetic acid (TFA – Solution A), with subsequent sonication consisting of three times 30 s pulses separated by 20 seconds The homogenized tissues were then centrifuged at 16,000 rpm for 20

37 minutes at 4°C and the supernatants were collected This process was repeated with the pellet leftover The extracted peptide mixture was analysed by RP-HPLC (PerkinElmar series 200 pump/autosampler, Flexar PDA detector and Chromera v3.2 software) The total collected peptides from the extractions were loaded on the HPLC Samples were separated and eluted with a protocol of 100% to 40% solution A at a flow rate of 1 mL/min over 20 min for the synthetic peptides and 60 min for the extracted peptide mixture Eluted compounds were detected at wavelengths of 210 nm and 280 nm Mobile phases used were solution A (0.1% TFA) and solution

B (0.1% TFA in acetonitrile) A total of 12 fractions were collected in 5 min intervals for further analysis by mass spectroscopy (MS) Control synthetic peptides were tested in RP-HPLC and observed to elute at 42.5% acetonitrile for GLDRYSFMGGI-NH2; 43.5% acetonitrile for GMPMLRL-

NH2; 42% acetonitrile for MRYFL-NH2; and 58.5% acetonitrile for RPGW-NH2 Five-minute fractions were lyophilised and resuspended in 1% formic acid for MS analysis

Mass spectrometry analysis (LC-MS/MS analysis) and protein identification

Resuspended HPLC fractions were analyzed by LC-MS/MS on a Shimadzu Prominance Nano HPLC (Japan) coupled to a Triple T of 5600 mass spectrometers (ABSCIEX, Canada) equipped with a nano electrospray ion source The protocol has been detailed elsewhere [25] Briefly, approximately 6 àL of each extract was injected and de-salted on the trap column before entering a nano HPLC column (Agilent Technologies, Australia) for mass spectrometry analysis The mass spectrometer acquired 500 ms full scan TOF-MS data followed by 20 by 50 ms full scan product ion data Full scan TOFMS data was acquired over the mass range 350-1800 and for product ion MS/MS 100-1800 Ions observed in the TOF-MS scan exceeding a threshold of 100 counts and a charge state of +2 to +5 were set to trigger the acquisition of product ion The data were acquired and processed using Analyst TF 1.5.1 software (ABSCIEX, Canada)

Fragmentation data was analyzed by PEAKS v6.0 (BSI, Canada) software Sequences of peptides were determined by comparing the fragmentation patterns with those predicted from the S glomerata transcriptomes Search parameters were as follows: no enzyme was used; variable modifications included methionine oxidation, conversion of glutamine/glutamate to pyroglutamic acid, deamidation of asparagine and peptide amidation Precursor mass error tolerance was set to 0.1 Da and a fragment ion mass error tolerance was set to 0.1 Da de novo sequencing, database search and characterising unspecific post-translational modifications (PTMs) were used to maximise the identifications; false discovery rate (FDR) was set to ≤ 1%, and

38 the individual peptide ion score [-10*Log(p)] was calculated accordingly, where p is the probability that the observed match is a random event

To investigate the possible roles of the neuropeptides identified in S glomerata reproduction, 28 neuropeptides were selected and synthesized by China Peptides Co Ltd for in vivo bioassays

Farmed S glomerata (grown out from wild spat) were purchased from a retail outlet on the Sunshine Coast They were held in seawater for at least 24 h before the assay Ten oysters examined showed a fully developed gonad, sex ratio was approximate 50% males and 50% females The oysters were relaxed by immersion in MgCl26H2O, 50g/L, for 4 h for the valves to open to enable the injection of peptides (40 àg/oyster) Synthesized peptides were pooled into nine different groups (Table 2.2) based on their derivation from precursors that often contains several bioactive peptides Forty àg of each bioactive peptide within one peptide treatment group were pooled into 10 àL of distilled water (volume designed for one injection) Distilled water was used as negative control, and 10 àL serotonin (5HT), 50mM, was used as positive control, since it has been shown to induce spawning in several bivalve species (Gibbons and Castagna, 1984, Matsutani and Nomura, 1987, Ram et al., 1993) Peptides were injected into the adductor muscle of S glomerata (n/group) with a Hamilton syringe (10R-GT syringe)

After each injection, oysters were placed individually into 540 mL plastic containers, which were randomly distributed to avoid any environmental influence Electric fans were used to ventilate the containers Spawning was confirmed by the observation of gametes, while fecundity was determined by sampling and counting eggs in each container with females 12 h post-injection using a stereoscope (Motic SMZ140-Fbled)

This bioassay was designed to investigate the involvement of APGWa and Buccalin at earlier development stages of the S glomerata gonad by implantation of cholesterol pellets (as implants) Each implant contained 50 àg (0.05mg) of peptide, either APGWa or buccalin and neuropeptide-free implants for negative control APGWa and Buccalin were chosen because they are both found in other oysters, where they function as a reproductive neuropeptides, however, their roles in SRO are still unknown Pellets were implanted into the adductor muscle using a chip

39 implant applicator (Trovan Ltd Germany) In total, three treatments, including blank (negative control; n$), APGWa (n$) and buccalin (n$) were tested S glomerata were obtained from PSFI To permit implantation, oysters were relaxed by immersion into MgCl26H20 (50g/L)

Fifty cholesterol implants (size: 1.5x 3 mm, weight: 5 mg) were made using a mixture containing 230 mg cholesterol + 15 àL copha + 10 mg peptides [(0.05 mg RPGWa + 0.05 mg KPGWa + 0.05 mg SPGWa + 0.05 mg APGWa)/each implant x 50 implants = 10 mg)] were used Each final implant weighed 5 mg and contained 0.2 mg of peptides Blank implants as a negative control were made using the mixture minus the peptides The same procedures were applied to make 50 buccalin implants that contain four distinct buccalin, 0.05 mg each (ALDRYSFFGGL-NH2, ALDKYGFFGGI-NH2,GLDRYSFMGGI-NH2, GLDRYSFMGGI-NH2 andGLDRYGFAGSL-NH2)

Results

Visceral ganglion RNA-seq provided 58 million nucleotide reads, that were assembled into 124,250 and 75,122 contigs, and 68,271 and 41,686 unigenes for male and female, respectively Gonad RNA-seq provided 37 million nucleotide reads, that were assembled into 100,109 and 85,668 contigs, and 47,033 and 37,54 unigenes for male and female, respectively From the unigene database, open reading frames were obtained then used for BLAST identification of mollusc neuropeptides

2.3.2 Neuropeptides identified from transcriptomes and LC-MS/MS analysis

In total, 28 neuropeptide precursors, including 11 putative full-length and 17 putative partial-length neuropeptide precursors were identified within the gender-specific visceral ganglia and gonad transcriptomes (Supplementary file S1) From these precursors, numerous bioactive neuropeptides were predicted to be released The majority of neuropeptides were identified from the visceral ganglia transcriptome (22 neuropeptides) and less from the gonad transcriptome (12 neuropeptides) (Table 2.1) A few neuropeptides were identified within only one sex Although the majority of previously known molluscan neuropeptides were identified, no ELH or GnRH peptide precursor was found by LC-MS/MS However, the ELH and GnRH precursors were identified within additional transcriptomic data, Ertl et al 2016, and a draft

S glomerata genome, respectively Several of the identified neuropeptides were synthesised for the in vivo bioassay (Table 2.1)

Table 2.1 Summary of neuropeptides deduced from the S glomerata transcriptomes

Note: shade means “positive” to each criterion on top of each column for each peptide on the left

To investigate the presence of neuropeptides, peptides were extracted from the visceral ganglia of female S glomerata and separated by RP-HPLC (Figure 2.2) Fractions were collected between 5-65 minutes for LC-MS/MS analysis Neuropeptides derived from 11 precursors were identified (Table 2.1), including APGWa, Buccalin, FCAP, FMRFamide, FXRLamide, GGN, Myoinhititin, Myomudulin 2, Rxlamide, SCAP and Tachykinin

Na m e o f pe pt ide s Pe pt ide s f or b io as sa y Pr ec ur so r f ul l l eng th? Si gna l pe pt ide (a a) Pr edi ct ed am ida tio n No o f bi oa ct iv e pe pt ide s Le ng th o f pr ec ur so rs (a a) Co nf irm ed b y M S Go na d Vi sc er al Ga ngl ia

No Ma le Fe m al e Ma le Fe m al e

Figure 2.2 RP-HPLC chromatogram and identification of neuropeptides extracted from visceral ganglia of female S glomerata (A) RP-HPLC chromatogram (B) S glomerata APGWa precursor shows peptide fragments identified by MS/MS, including which fraction (numbers on the scale bar) that were present (C) S glomerata Buccalin precursor shows peptide fragments identified by MS/MS, including which fractions (numbers on the scale bar) they were present Signal peptides in yellow, bioactive peptides in grey, amidation sites in aqua and cleavage sites (KR) in red The peak shows the concentration of peptides running through the HPLC apparatus at a time

The S glomerata ELH (Sg-ELH) transcript encodes a 169-residue precursor protein that is likely cleaved to produce two separate bioactive ELH peptides: ELH1 and ELH2 (Table 2.1, Figure 2.3) An alignment with other known oyster ELH shows that the Sg-ELH precursor shares more similarity with C gigas ELH (Cg-ELH; 64.4%) than P fucata ELH (Pf-ELH; 40.0%) High identity is

44 found within the bioactive neuropeptide regions (Figure 2.3B); 88.8% with Cg-ELH1, 2 and 53.8% with Pf-ELH1,2

Figure 2.3 Identification and characterization of S glomerata ELH precursor in comparison with other oysters (A) Amino acid sequences: signal peptides in yellow, bioactive peptides in grey where ELH1 is located before ELH2, amidation sites in aqua and cleavage sites in red (B) Alignment of ELH1 and ELH2 sequences: blue shading represents conservation of amino acid (C) Schematic diagrams illustrating the organisation of ELH precursors, (C gigas and P fucata precursor sequences were obtained from a supplementary file provided by

A precursor of S glomerata GnRH (Sg-GnRH) was identified that contains an N-terminal signal peptide, one bioactive GnRH peptide, and a GnRH associated peptide (Figure 2.4) Sg-GnRH shares high similarity within the bioactive GnRH peptide with other oysters (Figure 2.4B); it is identical between Sg- GnRH and Cg-GnRH, while there is only one amino acid difference with Pf- GnRH

Figure 2.4 Identification and characterization of S glomerata GnRH precursor in comparison with other oysters (A) Amino acid sequences: signal peptides in yellow, bioactive peptides in grey, amidation sites in aqua and cleavage sites in red; (B) Alignment of bioactive sequences: blue shading represents conservation of amino acid; (C) Schematic diagrams illustrating the organisation of precursors, (C gigas and P fucata precursor sequences were obtained from a supplementary file provided by Stewart et al (2014))

A full-length S glomerata APGWa (Sg-APGWa) was identified that contains 250 amino acids and is predicted to be cleaved to release 10 copies of four different bioactive peptides, including the amidated tetrapeptides: 1) APGWa (x3), 2) KPGWa (x3), 3) SPGWa (x1) and 4) RPGWa (x3) (Figure 2.5) The precursor Sg-APGWa is most conserved with other oyster APGWa precursors in the tetrapeptide regions (>90% identity with Cg-APGWa and Pf-APGWa) Sg- APGWa was confirmed by MS with two sequences (IKSFVDKRRP and RAPGWGKRSEMEKR) that were detected in fraction 2 (Figure 2.5A)

Figure 2.5 Identification and characterization of S glomerata APGWa precursor in comparison with other oysters (A) Amino acid sequences: signal peptides in yellow, cleavage sites in red and bioactive peptides in grey, amidation sites in aqua; S glomerata APGWa precursor shows peptide fragments identified by MS/MS, including the fraction (numbers on the scale bar) they were present (B) Schematic diagrams illustrating the organisation of precursors, (*) Peptides synthesized for bioassay APGWa precursors of C gigas and P fucata were obtained from a supplementary file provided by Stewart et al (2014)

A full-length buccalin precursor (Sg-buccalin) was identified that consists of 265 amino acids, and predicted to be processed to release nine distinct bioactive amidated peptides: 1) ALDRYSFFGGLa, 2) ALDKYGFFGGIa, 3) ALDRYGFAGSLa, 4) GLDRYNFFGGIa, 5) GLDRYGFAGSLa, 6) GLDRYSFMGGIa, 7) KLDRFGFMGGLa and 8) LDSHRFFGGLa and 9) RLDSHRFFGGLa (Figure 2.6) Its precursor also shows high identity with other oysters only within bioactive regions, and with overall identity of 80% and 84% with Cg-buccalin Cr-buccalin, respectively MS peptides matched to transcriptomic Sg-buccalin in three segments, including RYGFAGSLGKR (fraction 2 and 4),

RYGFAGSLGKRfALDRYGFIGSLGKR (fraction 2), and GKRRLDSHRFFGGLGKRAADQYENQG (fraction

Figure 2.6 Identification and characterization of S glomerata buccalin precursor in comparison with other oysters (A) Amino acid sequence: signal peptides in yellow, bioactive peptides in grey, amidation sites in aqua and cleavage sites in red S glomerata Buccalin precursor shows peptide fragments identified by MS/MS, including which fractions (numbers on the scale bar) they were present; (B) Schematic diagrams illustrating the organisation of precursors (*) Peptide synthesized for bioassays Buccalin precursor sequences of C gigas and

C rhizophorae oyster were obtained from a supplementary file provided by Stewart et al (2014)

S glomerata neuropeptide precursor proteins such as, allatotropin, conopressin, GGN,

GPA2, LFRYamide, NPY and PKYMDT were also identified While the PKYMDT precursors contain a single putative bioactive peptide, others such as the CCAP, LASGLVamide, LFRFamide, LRNFVamide and pedal precursor contain at least two predicted bioactive peptides (Table 2.1) Alignment of the S glomerata neuropeptide precursors with other mollusc species (e.g C gigas,

P fucata, Aplysia californica and Lottia gigantea) confirms the identification of allatotropin,

CCAP, LFRFa and LRNFVa (Figure 2.7)

Figure 2.7 Identification and characterization of other S glomerata neuropeptides Schematic diagrams show the organisation of neuropeptide precursors and multiple sequence alignment of bioactive peptide between mollusc species Blue shading represents conservation of amino acid Precursor sequences of neuropeptide of other molluscs were obtained from a supplementary file provided by Stewart et al

Discussion

In this study, we have undertaken transcriptome sequencing from male and female of S glomerata, which provided a database of over 100,000 unigenes from visceral ganglion and over

84,000 unigenes from gonads This enabled the identification of 28 putative neuropeptide precursors that are likely to be proteolytically processed to release numerous bioactive neuropeptides Thereafter, we have investigated their potential role in the regulation of S glomerata reproduction

Allatotropin was initially found in Manduca sexta as a stimulant for generating juvenile hormone (Kataoka et al., 1989), however, the role of Allatotropin in oysters is still unknown In this study, treatment of 40 àg per individual peptide per oyster in combination with LASGLVa did not induce female oysters to spawn

Although ELH and GnRH are abundant in molluscan neural tissues e.g visceral, cerebral or pedal ganglia (Brown and Mayeri, 1989, Morishita et al., 2010), they may also be present within other tissues such as mantle, gills, adductor muscle and gonad, or even the hemolymph (Bigot et al., 2012, Treen et al., 2012) In this study, we did target the visceral ganglia and gonad, yet the Sg-ELH and Sg-GnRH were not detected Their absence in these tissues could be explained by either no or low levels of expression S glomerata also contains another major ganglion, the cerebral ganglia, which was not analysed since the proximity of the visceral ganglia with the gonad was assumed to be most appropriate for neuroendocrine signalling Fortunately, a transcriptome generated from mixed S glomerata tissues (hemolymph, mantle, gill, gonad, digestive tissue and adductor muscle) enabled their identification, although it is unclear which tissue and what sex contained the transcripts ELH is known to be involved in the reproduction of gastropods (Li et al., 1999, Morishita et al., 2010), were it induces egg laying behaviour in sexually mature Aplysia (Nagle et al., 1988, Nambu and Scheller, 1986, Scheller et al., 1983) However, there is no report on function of ELH as an oyster spawn inducer As a first step towards understanding its potential role in oyster reproduction, we first analysed the Sg-ELH precursor primary sequence, showing that its organization is similar to that of other oysters, where it contains two ELH-like peptides (ELH1 and ELH2) within the same precursor Their high level of identity with respective Cg-ELH1 and ELH2 confirms a critical regulatory role in S glomerata

Moreover, our spawning bioassay (Bioassay 1) did show that ELH1 and ELH2 could induce spawning in 70% of the females within 12 h post-peptide injection This result provides a first insight into the role of ELH in an oyster, as an egg-laying (spawning) hormone

GnRH is a well-studied reproductive neuropeptide in vertebrates and there has been accumulating evidence for GnRH peptides in mollusc species (Tsai and Zhang, 2008) In bivalves, GnRH peptides were found in visceral ganglia of both sexes in the scallop P yessoensis and C gigas (Bigot et al., 2012, Stewart et al., 2014, Treen et al., 2012) GnRH tends to be strictly conserved within the bioactive GnRH region, even between invertebrates and chordates (Stewart et al., 2014, Tsai, 2006) In this study, we found that the Sg-GnRH peptide was identical to the Cg-GnRH peptide The GnRH peptide induced 86% of mature S glomerata females to spawn Its direct role may be to regulate gamete proliferation or maturation, as has been demonstrated in P yessoensis (Treen, 2012), where in vitro application of mammalian GnRH could stimulate spermatogonial proliferation Also, an extract of scallop cerebral and pedal ganglia had the same effect (Nakamura et al., 2007) The presence of GnRH receptors in mature gonads of M edulis (Pazos and Mathieu, 1999) and C gigas (Morishita et al., 2010) also provides evidence for a reproductive role in these species Our preliminary analysis of the S glomerata gonads did not find a GnRH receptor

The APGWamide was first detected from the ganglia of the gastropod Fusinus ferrugineus (Kuroki et al., 1990) and later in other gastropods e.g Lymnaea stagnalis, A californica and L gigantea (Fan et al., 1997, Smit et al., 1992, Veenstra, 2010) and oysters e.g C gigas (Bernay et al., 2006, Stewart et al., 2014) Similar to the C gigas, we did find the Sg- APGWa in the visceral ganglia APGWa precursors tend to be conserved among mollusc species in both organization and number of bioactive peptides (approximately 8-10 repeats) (Fan et al., 1997, Smit et al.,

1992, Stewart et al., 2014, Veenstra, 2010, York et al., 2012) However, we found that the Sg- APGWa precursor has greater diversity of the types of tetrapeptides than C gigas and P fucata (i.e four of each APGWa, KPGWa, SPGW and RPGWa in S glomerata vs three of each APGWa, KPGWa, and RPGWa) Meanwhile, gastropods typically have only one type of tetrapeptide in their precursors, such as APGWa in Haliotis asinina (York et al., 2012), L gigantea (Veenstra,

2010) and A californica (Fan et al., 1997) Conservation of Sg-APGWa within the bioactive regions was high, showing 90% identity within C gigas or P fucata This similarity suggests a similar role for APGWa in oysters

APGWa’s role in reproduction is well established in molluscs, including for example its involvement as a male reproductive stimulant through activation of genital eversion in L stagnalis (Koene, 2010) Also, it can activate spermiation in Helix aspersa and induce male spawning in H asinina (Chansela et al., 2008) In oyster, APGWa regulates egg transportation and spawning in female C gigas (Bernay et al., 2006) Moreover, APGWa present in sperm may act as a pheromone to trigger female spawning when females come into contact with this sperm (Bernay et al., 2006) In this study, we confirm a role for APGWa in regulating female oyster spawning, showing 100% bioactivity (Bioassay 1)

APGWa seems to contribute to the development of gonad condition at early maturation stage APGWa appears to not only induce spawning in S glomerata, but also stimulates S glomerata gonad development and maturation In this study, APGWa-treated oysters obtained a significant higher gonad condition index and fertilisation rate than the control (peptide-free implantation, P70%) both in buccalin and APGWa treatments in the implantation bioassay (more than 70% females in buccalin and APGWa vs about 50% in Control) warrants further investigation into a potential role in sex change in oysters

Conclusions

Transcriptome analysis of S glomerata has revealed 28 neuropeptide precursors, from which 11 neuropeptides could be confirmed by peptidomics MS analysis within the visceral ganglia The synthesized ELH, GnRH, APGWa, buccalin, CCAP and LFRFa could induce spawning in ripe wild-caught oysters APGWa and buccalin enhanced gonad development, and increased the efficiency of S glomerata gamete fertilisation This is the first report on neuropeptides identified from S glomerata transcriptomes that have a regulatory role in oyster reproduction.

Acknowledgements

We gratefully acknowledge the funding support of the Australian Centre for International Agricultural Research (ACIAR), John Allwright PhD fellowship to V.V.I, the Australian Seafood CRC and FRDC (Project No.2012/713) to AE and WO, and the University of the Sunshine Coast (USC) Also, we thank the Australian Research Council (SFC) We thank PSFI, NSW for generously providing samples and access to facilities for spawning and implantation bioassays Special thanks to Steve O’Connor, Michael Dove, Kyle Johnston, Brandt Archer (PSFI, NSW), and Brad Harding and Bronwyn Rotgans (USC) for their support We gratefully thank Dr Alun Jones (Institute for Molecular Bioscience, the University of Queensland) for advice and assistance with the LC-MS/MS.

Supplementary data

Supplementary file S1 Amino acid sequences of Saccostrea glomerata and other molluscan neuropeptides used for schematic diagrams and alignments.

References

BERNAY, B., BAUDY-FLOC'H, M., ZANUTTINI, B., ZATYLNY, C., POUVREAU, S & HENRY, J 2006 Ovarian and sperm regulatory peptides regulate ovulation in the oyster Crassostrea gigas Molecular Reproduction and Development, 73, 607-616

BIGOT, L., ZATYLNY-GAUDIN, C., RODET, F., BERNAY, B., BOUDRY, P & FAVREL, P 2012 Characterization of GnRH-related peptides from the Pacific oyster Crassostrea gigas Peptides, 34, 303-310 BROWN, R & MAYERI, E 1989 Positive feedback by autoexcitatory neuropeptides in neuroendocrine bag cells of Aplysia The Journal of Neuroscience, 9, 1443-1451

BRUNAK, S., ENGELBRECHT, J & KNUDSEN, S 1991 Prediction of human mRNA donor and acceptor sites from the DNA sequence Journal of molecular biology, 220, 49-65

BURBACH, J P H 2011 What are neuropeptides? Neuropeptides: Methods and protocols New York,

CHANSELA, P., SAITONGDEE, P., STEWART, P., SOONKLANG, N., STEWART, M., SUPHAMUNGMEE, W.,

POOMTONG, T & SOBHON, P 2008 Existence of APGWamide in the testis and its induction of spermiation in Haliotis asinina Linnaeus Aquaculture, 279, 142-149

CONN, P J & KACZMAREK, L K 1989 The bag cell neurons of Aplysia A model for the study of the molecular mechanisms involved in the control of prolonged animal behaviors Mol Neurobiol, 3, 237-73

CONZELMANN, M., WILLIAMS, E A., KRUG, K., FRANZ-WACHTEL, M., MACEK, B & JÉKELY, G 2013 The neuropeptide complement of the marine annelid Platynereis dumerilii BMC genomics, 14, 906 CROPPER, E C., MILLER, M W., TENENBAUM, R., KOLKS, M., KUPFERMANN, I & WEISS, K R 1988

Structure and action of buccalin: a modulatory neuropeptide localized to an identified small cardioactive peptide-containing cholinergic motor neuron of Aplysia californica Proceedings of the National Academy of Sciences, 85, 6177-6181

CUMMINS, S F., TOLLENAERE, A., DEGNAN, B M & CROLL, R P 2011 Molecular analysis of two

FMRFamide encoding transcripts expressed during the development of the tropical abalone

Haliotis asinina Journal of Comparative Neurology, 519, 2043-2059

DINAMANI, P 1974 Reproductive cycle and gonadial changes in the New Zealand Rock Oyster,

Crassostrea glomerata New Zealand Journal of Marine and Freshwater Research, 8, 39-65

DOVE, M C & O'CONNOR, W A 2009 Commercial assessment of growth and mortality of fifth generation Sydney Rock Oysters, Saccostrea glomerata (Gould, 1850) selectively bred for faster growth Aquaculture Research, 40, 1439-1450

FAN, X., CROLL, R P., WU, B., FANG, L., SHEN, Q., PAINTER, S D & NAGLE, G T 1997 Molecular cloning of a cDNA encoding the neuropeptides APGWamide and cerebral peptide 1: Localization of APGWamide like immunoreactivity in the central nervous system and male reproductive organs of Aplysia Journal of Comparative Neurology, 387, 53-62

FRICKER, L D Neuropeptides and other bioactive peptides: from discovery to function Colloquium Series on Neuropeptides 2012: Morgan & Claypool Life Sciences, 1-122

GIBBONS, M & CASTAGNA, M 1984 Serotonin as an inducer of spawning in six bivalve species

KATAOKA, H., TOSCHI, A., LI, J P., CARNEY, R L., SCHOOLEY, D A & KRAMER, S J 1989 Identification of an allatotropin from adult Manduca sexta Science, 243, 1481-1483

KOENE, J M 2010 Neuro-endocrine control of reproduction in hermaphroditic freshwater snails: mechanisms and evolution Frontiers in behavioral neuroscience, 4

KUROKI, Y., KANDA, T., KUBOTA, I., FUJISAWA, Y., IKEDA, T., MIURA, A., MINAMITAKE, Y & MUNEOKA, Y

1990 A molluscan neuropeptide related to the crustacean hormone, RPCH Biochemical and

LAWRENCE, D & SCOTT, G 1982 The determination and use of condition index of oysters Estuaries, 5,

LI, L., GARDEN, R W., FLOYD, P D., MOROZ, T P., GLEESON, J M., SWEEDLER, J V., PASA-TOLIC, L &

SMITH, R D 1999 Egg-laying hormone peptides in the Aplysiidae family Journal of experimental biology, 202, 2961-2973

MATSUTANI, T & NOMURA, T 1987 In vitro effects of serotonin and prostaglandins on release of eggs from the ovary of the scallop, Patinopecten yessoensis General and Comparative Endocrinology,

MILLER, M W., BEUSHAUSEN, S., CROPPER, E C., EISINGER, K., STAMM, S., VILIM, F S., VITEK, A., ZAJC,

A., KUPFERMANN, I & BROSIUS, J 1993 The buccalin-related neuropeptides: isolation and characterization of an Aplysia cDNA clone encoding a family of peptide cotransmitters J Neurosci,

MORISHITA, F., FURUKAWA, Y., MATSUSHIMA, O & MINAKATA, H 2010 Regulatory actions of neuropeptides and peptide hormones on the reproduction of molluscs The present review is one

56 of a series of occasional review articles that have been invited by the Editors and will feature the broad range of disciplines and expertise represented in our Editorial Advisory Board Canadian

NAGLE, G T., PAINTER, S., BLANKENSHIP, J & KUROSKY, A 1988 Proteolytic processing of egg-laying hormone-related precursors in Aplysia Identification of peptide regions critical for biological activity Journal of Biological Chemistry, 263, 9223-9237

NAKAMURA, S., OSADA, M & KIJIMA, A 2007 Involvement of GnRH neuron in the spermatogonial proliferation of the scallop, Patinopecten yessoensiss Molecular reproduction and development,

NAMBU, J R & SCHELLER, R H 1986 Egg-laying hormone genes of Aplysia: evolution of the ELH gene family The Journal of Neuroscience, 6, 2026-2036

NELL, J A 2006 Manual for mass selection of Sydney rock oysters for50 fast growth and disease resistance Fisheries Research Report series 13 NSW: NSW Department of Primary

NUURAI, P., POLJAROEN, J., TINIKUL, Y., CUMMINS, S., SRETARUGSA, P., HANNA, P., WANICHANON, C &

SOBHON, P 2010 The existence of gonadotropin-releasing hormone-like peptides in the neural ganglia and ovary of the abalone, Haliotis asinina Acta Histochemica, 112, 557-566

O'CONNOR, W A., DOVE, M C & FINN, B 2008a Sydney Rock Oysters: Overcoming constraints to commercial scale hatchery and nursery production NSW

O'CONNOR, W A., DOVE, M C., FINN, B & O’CONNOR, S J 2008b Manual for hatchery production of sydney rock oysters, Saccostrea glomerata IDP NSW NSW Department of Primary Industries Port

Stephens Fisheries Center Taylors Beach NSW 2316, Australia http://www.dpi.nsw.gov.au/research/areas/aquaculture/outputs/2008/oconnor3

PAZOS, A & MATHIEU, M 1999 Effects of five natural gonadotropin-releasing hormones on cell suspensions of marine bivalve gonad: stimulation of gonial DNA synthesis General and

RAM, J L., CRAWFORD, G W., WALKER, J U., MOJARES, J J., PATEL, N., FONG, P P & KYOZUKA, K 1993

Spawning in the zebra mussel (Dreissena polymorpha): activation by internal or external application of serotonin Journal of Experimental Zoology, 265, 587-598

SCHELLER, R H., JACKSON, J F., MCALLISTER, L B., ROTHMAN, B S., MAYERI, E & AXEL, R 1983 A single gene encodes multiple neuropeptides mediating a stereotyped behaviour Cell, 32, 7-22

SMIT, A., JIMÉNEZ, C., DIRKS, R., CROLL, R & GERAERTS, W 1992 Characterization of a cDNA clone encoding multiple copies of the neuropeptide APGWamide in the mollusk Lymnaea stagnalis The

SPSS STATISTICS 22 2013 IBM SPSS statistics 22 core system user's guide

STEWART, M J., FAVREL, P., ROTGANS, B., WANG, T., ZHAO, M., SOHAIL, M., O'CONNOR, W A., ELIZUR,

A., HENRY, J & CUMMINS, S F 2014 Neuropeptides encoded by the genomes of the Akoya pearl oyster Pinctata fucata and Pacific oyster Crassostrea gigas: a bioinformatic and peptidomic survey BMC genomics, 15, 840

STRUMWASSER, F., SCHILLER, D & KENT, S Synthetic neuropeptide egg-laying hormone (ELH) of Aplysia californica induces normal egg-laying: structure-activity studies Soc Neurosci Abstr 1987, 38

TAMURA, K., STECHER, G., PETERSON, D., FILIPSKI, A & KUMAR, S 2013 MEGA6: molecular evolutionary genetics analysis version 6.0 Molecular biology and evolution, 30, 2725-2729

TREEN, N., ITOH, N., MIURA, H., KIKUCHI, I., UEDA, T., TAKAHASHI, K G., UBUKA, T., YAMAMOTO, K.,

SHARP, P J & TSUTSUI, K 2012 Mollusc gonadotropin-releasing hormone directly regulates gonadal functions: A primitive endocrine system controlling reproduction General and

TRENAMAN, R., LIVINGSTONE, S & CREESE, A 2015 Aquaculture Production Report, NSW DPI, 10pp

TSAI, P.-S 2006 Gonadotropin-releasing hormone in invertebrates: Structure, function, and evolution

TSAI, P.-S & ZHANG, L 2008 The emergence and loss of gonadotropin-releasing hormone in protostomes: orthology, phylogeny, structure, and function Biology of reproduction, 79, 798-805 VEENSTRA, J A 2010 Neurohormones and neuropeptides encoded by the genome of Lottia gigantea, with reference to other mollusks and insects General and Comparative Endocrinology, 167, 86-

YORK, P S., CUMMINS, S F., DEGNAN, S M., WOODCROFT, B J & DEGNAN, B M 2012 Marked changes in neuropeptide expression accompany broadcast spawnings in the gastropod, Haliotis asinina

Can genetic diversity be maintained across multiple mass selection lines of

Introduction

Sydney rock oyster, Saccostrea glomerata is an important aquaculture species in New South Wales (NSW) and Southern Queensland, Australia, with NSW producing 7,793,390 dozen oysters with a farm gate value of $31.8 million for 2013/2014 (Trenaman et al., 2014)

Since 1990, in NSW Australia, various different mass selection programs were conducted to improve growth and, independently, to prove survival in S glomerata for resistance to “winter mortality” syndrome (unknown causative agent) After five generations of mass selection for fast growth, the S glomerata lines at Port Stephens, NSW, could reach market size (i.e 50g whole weight) 10-11 months earlier than control unselected lines (Dove and O'Connor, 2009, Nell and Perkins, 2005) Similarly, by the third generation of mass selection, loss to winter mortality at Quibray Bay was reduced to 22% vs 46% for the unselected lines The winter mortality resistant line was additionally selected for fast growth and showed similar improvements in growth to that of the Port Stephens lines

Typical for almost any genetic program, it was a requirement in mass selection for oysters to “close” the population, and stop adding wild animals to the breeding nucleus Wild animals are genetically unimproved and may diminish genetic selection response of the line However, there are examples in S glomerata and C gigas in Australia, and in shrimp, where closing the

60 nucleus of genetic programs and conducting selection over generations without monitoring the pedigree may be accompanied by inbreeding, especially when there are high selection intensities e.g for fast growth or disease resistance (Appleyard and Ward, 2006, Nguyen, 2009) The genetic diversity of broodstocks for breeding can be restricted due to selection of just a few outstanding families, or simply due to chance sampling of a few families where the species are highly fecund, and where one or two females can contribute the majority of the next generation (Appleyard and Ward, 2006, Boudry, 2008, Taris et al., 2006) Oysters may be particularly vulnerable to loss of families and consequent inbreeding when no pedigree is maintained as a single oyster female can release millions of eggs in one spawning event (O'Connor et al., 2008b)

While mass selection can improve given traits quickly, the selection of individuals on the basis of best performance of certain traits without regard to pedigree information can result in inbreeding and loss of potentially valuable alleles and net additive genetic variation Inbreeding can adversely affect genetic breeding programs in a number of ways The rate of selection response is expected to slow over generations in inbred lines (Bentsen and Olesen, 2002) and mating with close relatives unmasks recessive deleterious alleles resulting in inbreeding depression (Amos and Balmford, 2001, Amos and Harwood, 1998) In hatchery populations, inbreeding often causes negative effects on stock performance and production traits (Evans et al., 2004b, Kincaid, 1983, Launey et al., 2001) For example, there was high inbreeding depression for yield after selection for individual growth and survival in C gigas (Evans et al., 2004b)

Mass selection without pedigree information results in the dilemma of how to close the nucleus without inbreeding Recently Knibb et al (2014) demonstrated a possible solution to this dilemma (maintaining genetic diversity while trying to remain high response to selection of an interested trait), by showing that genetic variation, though lost in different individual selection lines of banana shrimp, Fenneropenaeus merguiensis, is well maintained between them, so interline crossing can help restore the genetic diversity among selected lines even after many generations of selection These observations were made in banana shrimp, but have not been reported generally elsewhere, and it is unknown whether this principle applies generally, or specifically to the S glomerata mass selection lines Here we test this principle that multiple mass selection lines preserve genetic variation by estimating the levels of genetic variation (assessed using both mtDNA haplotypes and 10 DNA microsatellite loci) in various long term mass selection lines of S glomerata, including those selected over five and seven generations for disease resistance and fast growth, respectively We also test the efficiency of the approach of

61 subdivided lines in maintaining diversity, by assessing if the net genetic diversity maintained among lines is not statistically significantly different from that in samples from a wild population (P>0.05)

However, this paper does not test the performance of the different lines in terms of growth, reproduction, survival, nor does attempt to correlate the genetic diversity data with line performance.

Materials and methods

Lines selected either for fast growth or winter mortality syndrome resistance (WM) all originated from the same base population, except line A which was established from wild winter mortality survivors from Shoalhaven River and Merimbula Lake, NSW (Figure 3.1) The base population was formed by interbreeding using equal numbers of oysters from the four key commercial farming estuaries (Wallis Lake, Hawkesbury River, Port Stephens and Georges River) in February 1990 (Nell, 2006) Then, the base population was used to establish breeding sub- lines in Port Stephens and Georges River The first generation of fast growth S glomerata from the Port Stephens sub-line was produced in 1993 and selected over seven generations for growth rate until 2009 (two of these sublines were designated L1 and L2, Figure 3.1) Another sub-line from Georges River lines (line B, Figure 3.1) was selected for two generations for fast growth in the Georges River from 1992 to 1994, then selected for WM resistance at Quibray Bay (Georges River) from 1 st generation in 1997 to 5 th generation in 2008 A fourth sub-line, (A, Figure 3.1) was created in 2006 and selected for fast growth

The original base population that subsequently formed lines A, B, L1, L2 was generated from a series of eight separate mass spawning and fertilization events in February, 1990 Each spawning event used 20 males and 80 females comprising approximately 10 males and 15 females from each of the four different commercial lines Equal numbers of fertilized eggs from each of the eight spawning events were stocked in several 20,000L tanks In August 1990, ensuring spat were split randomly in four breeding lines, initially designated to be kept separately, for the Georges River (one of the four lines, namely line B, was used for this study)

62 and four more, also initially designated to be kept separately, for Port Stephens (two of these, namely L1 and L2, were used for this study)

When disease occurred in 1997, the Georges River lines were reorganised to form three disease resistant lines (Nell, 2006), one of these was used for this study (Line B, Figure 3.1) Approximately 40 males and 90 females from line B were split into four separate spawning events or fertilisations to produce the next generation Oysters were spawned naturally through temperature and salinity manipulation or strip spawned Fertilisation was controlled by separating spawning males and females into their own containers before a series of batches were fertilised Once fertilised, the groups were pooled so that each female contributed a similar number of fertilised eggs to the pooled batch (Nell, 2006) Similar procedures were used to produce line A from a wild base population of winter mortality survivors from the Shoalhaven River and Merimbula Lake, NSW (Figure 3.1)

Figure 3.1 Diagrammatic representation of various S glomerata lines

Port Stephens fast growth lines Georges River fast growth breeding lines

Each arrow represents one generation

For validation of Mendelian inheritance of the microsatellite DNA markers, mantle tissues from 10 sires and 10 dams comprising 10 single pair matings and their full sib progeny (approximately 10 full sib offspring per family) were sampled

For the assessment of variation among lines, up to 120 pieces of mantle tissue were sampled from each of the four selected lines (B – generation 5, A – generation 3, L1 – generation

7, L2 – generation 7) when they were being reproduced at the Port Stephens Fisheries Institute, NSW Thirty samples were also taken from wild animals (W) at the Port Stephens Fisheries Institute in Cromarty Bay All samples were taken in April 2012, and they were preserved in 70% ethanol and shipped to the University of the Sunshine Coast (USC), where they were stored at -

20 o C The minimum sample size for analysis of DNA microsatellite loci and mitochondrial Cytochrome C oxidase subunit I (COX1) was 26 and 15 individuals for each line, respectively Each list of samples is a random one by order, and so truncating the list to a common number for all lines serves functionally as randomly sampling

DNA extraction was based on the NaCl extraction protocol (Lopera-Barrero et al., 2008) The integrity of the DNA was verified by horizontal electrophoresis in a 0.9% agarose gel, at 100 volts for 45 minutes in a 0.6xTBE buffer (500 mM Tris-HCl, 60 mM boric acid, and 83 mM EDTA) The gel was dyed with ethidium bromide, verified and captured in GeneSnap with the Syngene System (Bio-Rad) In addition, the quality and quantity of the DNA were then evaluated using a Nanodrop 2000 (Thermo Scientific) at the absorbance of 260/280nm Good DNA templates were then diluted in molecular grade water (Amresco) to 25 ng àL -1

3.2.5 Mitochondrial DNA sequencing and analysis

The mitochondrial cytochrome C oxidase subunit I (COX1) was amplified using universal primers (LCO1490 and HCO2198) developed by Folmer et al (1994) in 25 àL reactions using MyTaq DNA polymerase (Bioline) Nity six DNA samples from five different stocks (19-20 samples from each population) were used for COX1 amplification and sequencing using an ABI 3730XL DNA analyser Sequencher 5.0 (Gene Codes Corporation, 2011) was used to trim and correct

64 errors in COX1 sequences Mega 5.0 (Tamura et al., 2011) was used for sequence alignment, identification of haplotypes and haplotype comparison

3.2.6 Validation and development of microsatellite primers

S glomerata DNA microsatellites developed from a transcriptome: Total RNA was extracted from ganglion and gonad tissues following the protocol for isolation of RNA using Trisure (Bio-38032; Bioline), which was then sent to BGI, Hong Kong, for transcriptome sequencing BGI conducted RNA-Seq with 2 lanes of 100 paired-end sequencing on the illumine HiSeq2000 Gonad and muscle tissues collected from individuals representing males and females from different developmental stages were kept in RNA later solution (Ambion) at -80 o C until required Microsatellites and their primers were detected utilising the acquired transcriptome in Nomachine software for Windows 4.4.1.7 Of those, 48 pairs of primers for potential microsatellites were selected and amplified in 12.5 àL PCR reactions containing 1xMyTaq reaction buffer, 0.125 mg ml -1 of bovine serum albumin, 30ng of template DNA, 0.02 àM of forward and reverse primers in which forward primers labelled with fluorescent dyes (FAM, NED, PET or VIC), and 1 unit of MyTaq DNA polymerase (Bioline) The reaction was amplified in an Eppendorf Mastercycler gradient thermal cycler using following steps: an initial denaturing for 5 min at 95 o C followed by 30 cycles of 95 o C for 30s, annealing for 15s, extension at 72 o C for 10s and final extension at 72 o C for 2 mins PCR products were qualified on 3% agarose gels run for 2 h at 140 volts to detect levels of polymorphism (high variation of alleles on a locus) and quality of amplified products On the basis of potential polymorphisms evident from agarose gel electrophoresis, 10 putatively polymorphic microsatellites were tested on eight different individual samples using fluorescent labelled primers and genotyped in an AB 3500 Genetic Analyser, Hitachi Seven of the ten were chosen for this study based on good amplification, repeatability and reliability of scoring and degree of allelic variation

S glomerata DNA microsatellites from previous publications: nine pairs of public domain microsatellite primers, developed by Banks et al (2006) and three by Queensland University of Technology (Nguyen, 2009) were tested on eight oyster samples using fluorescent labelled primers Three out of 12 (Sgo09, Sgo16 and Sgo21) were found to be suitable (reliably scored) using the same selection procedures as described above

Null allele and polymorphic information content (PIC) were analysed on Microchecker 2.2.0.3 (Van Oosterhout et al., 2004) and Cervus 3.0 (Marshall et al., 1998), respectively Number

65 of alleles, expected and observed heterozygosity, fixation index (Fis) based on 128 samples were analysed using Genalex 6.5 (Peakall and Smouse, 2012)

To determine whether the alleles at each of these ten loci are transmitted in a Mendelian fashion, we genotyped 10 single pair mated families which included the two parents and 10 to

11 offspring per family Out of 20 parents and 10 loci, there were 6 cases out of 200 (considering

10 loci) where we needed to invoke a null allele in order to explain the offspring genotypes by Mendelian inheritance from parents (Supplementary file S2_Table 2) So, these data (Supplementary file S2_Table 2) also allow an estimate of the proportion of null alleles If we just consider the 20 hatchery parents used to test for Mendelian Inheritance, then e.g at loci Sg21 we detected two different parents with possible nulls by considering their offspring, although it is possible further nulls were present, just not evident For our calculations, we make the assumption that if a null is detected at a given loci and given parental pair, then only one of the four possible parental alleles is a null (based simply on the likelihood that nulls are rare) Thus, for Sg21 we can calculate the frequency of nulls is 2 counts out of a total of 40 counts (20 individuals time 2 alleles), or 5% Similarly, we estimate null frequencies to be 0% for Sg9, 0% for Sg10, 0% for Sgo16, 5.0% for Sgo22, 0% for Sgo27, 2.5% for Sgo34, 2.5% for Sgo37, 0% for Sgo45 and 0% for Sgo46 Similarly, average null allele frequency per locus over all loci was estimated to be 1.5%

There was also one case of irregular inheritance that was most simply explained as mislabelling of family origin (i.e the irregular genotype fitted that expected for a different family)

Genotyping PCR products were conducted using an AB 3500 Genetic Analyser For preparation prior to genotyping, PCR products were diluted 10 to 50 fold depending on the yield of PCR amplified products 1 àL of each diluted PCR product was then transferred to 9 àL of a combination of Liz 600 (size standard V2.0) and HiDi formamide (7 àL Liz for 500 àL HiDi), then vortexed and briefly centrifuged before denaturation for 5 min at 95 o C, cooling of DNA samples for 2-3 mins in an ice bath in order to fix the DNA before transferring them to the genotyping machine Genotyping was conducted using the 500bp fragment analysis program

3.2.8 Data analysis and statistical methods

Output (.fsa files) from the genotyper were analysed and scored using GeneMarker 2.6.3 (Schmidt, 2011) Genotyping data was checked for null alleles using Microchecker 2.2.0.3 (Van Oosterhout et al., 2004) Mendelian inheritance patterns were tested with genotyping data from ten families and their offspring Genalex 6.5 (Peakall and Smouse, 2012) and Cervus 3.0 (Marshall et al., 1998) were used to analyse allele diversity, molecular variance, fixation index (Fis) and Hardy-Weinberg equilibrium test and polymorphic information content (PIC)

Colony 2.0.2.3 (Jones and Wang, 2010) was used to determine inbreeding and parentage assignment SPSS statistics 22 was used to conduct Chi-square tests (SPSS statistics 22, 2013).

Results

3.3.1 Microsatellite diversity within among lines and wild populations

Using 10 DNA microsatellite loci to analyse genetic diversity from four hatchery lines and one wild reference population we found over all the samples (but excluding those used for testing Mendelian inheritance) that the number of alleles ranged from eight to 37 alleles per microsatellite locus (raw data in Supplementary file S2_Table 3)

Restricting the data to a sample size of 26 per line to avoid the potential bias that more alleles will be found in larger samples, it was evident that the samples from the wild population tended to have a greater number (about twofold) of alleles at every locus compared with each individual selection line (Table 3.1), and these differences were statistically significant, except for line A (Figure 3.2)

However, if we combine the different alleles among the 104 samples from all four selection lines (Pooled hatch – 104), there are a total of 92 distinct alleles, which as a count is not statistically significantly different from the count for the wild (Figure 3.2, P>0.05) Even if we adjust the total sample size from all four selection lines to 26 by random sampling in order not to bias upwards the allele counts from lines with larger samples sizes (termed “Pooled hatch 26”), we still find the total number of alleles for the selection lines is 75, which is not significantly different from the 120 in the samples from the wild population (Table 3.1, Figure 3.2; P>0.05)

Full sib families (i.e individuals that shared the same sire and dam) were detected in each of the selected lines but not in the samples from the wild population (Table 3.2) B (WM resistant

67 line) and L1 (fast growth line) had the greatest number of full sib families, which concurred with them having the lowest Ne (Table 3.2) and fewest alleles (see previous section) Inbreeding was also higher in these two lines

Table 3.1 Number of alleles among loci in selected lines and wild samples

By running all the genotypes from the 4 selected lines and the one wild line together, we found no evidence of full sibs shared between the wild reference and any of the selected lines

Genetic relationships among individuals of five S glomerata lines or population were estimated using genetic distance matrices (Structure 2.2) The overall impression is that the hatchery lines tended to cluster apart but the samples from the wild population tended to overlap with the hatchery samples (Figure 3.3)

Table 3.2 Number of full sibs and effective population size among lines and wild samples

Lo ci ( nu m be r o f alle le s)

Total number of full sib families

Effective population size with upper and lower 95% confidence intervals

3.3.2 mtDNA haplotype diversity among populations

There were 13 different COX1 mtDNA haplotypes considering all four selected lines and the samples from the wild population (Table 3.3) The wild and hatchery samples shared the same two common haplotypes (i.e haplotypes A and B), except line B However, the hatchery lines also had five private haplotypes not shared with the wild, and vice versa for six haplotypes

(Table 3.3) The wild had substantially more mtDNA haplotypes than any of the selected lines and line B, which had the fewest DNA microsatellite alleles and also had the fewest haplotypes

Figure 3.2 Number of alleles and haplotypes among populations Bars with the same letter within a category (allele or haplotype) are not statistically significantly different assessed using Chi-square tests of counts (P>0.05) 1 : Number of alleles using 26 samples; 2 : Number of haplotypes using 15 samples; 3 : Number of alleles using

104 samples; 4 : Number of haplotypes using 60 samples

The correlation between total number of different DNA microsatellite alleles and the total number of different mtDNA haplotypes considering the four hatchery samples and the single wild sample was 0.880 (P0.05), whether we consider all the hatchery samples (i.e all 60) or just 15 samples randomly chosen from the four different hatchery lines in order to remove the potential bias of

N um be r o f a lle le N um be r o f h ap lo ty pe

Number of alleles Number of haplotypes

70 finding more haplotypes simply because we had more hatchery samples than those samples from the wild population (Figure 3.2)

The genetic distance between the wild and the lines or line vs line are small using Fst

Discussion

Seven new microsatellite markers for Sydney rock oyster, and three published ones were identified and found to be reliably and consistently scored using software (i.e GeneMarker) Alleles at these loci, for the most part, are transmitted and inherited in a Mendelian fashion Three of the loci were found to have low frequencies of null alleles Many other studies using microsatellite markers have also reported null alleles, particularly in studies of mollusc species (Astanei et al., 2005, Carlsson et al., 2006, Li et al., 2003, Reece et al., 2004) High frequencies of null alleles may produce pseudo-homozygotes and scoring errors, however in this study, the frequencies of null alleles that could be detected using parent-offspring data was low (2.1% overall), so null alleles would not affect the overall findings of this study

Table 3.3 Number of mtDNA haplotypes in selected lines and wild samples

Ha pl ot yp es

‡: Number of haplotypes estimated using 15 samples †: Number of haplotypes estimated using 60 samples

Breeding practices for S glomerata in NSW attempted to avoid loss of genetic diversity and inbreeding in the hatchery selection lines through the use of moderate number of parents for each generation (130 contributing parents per generation) Even so, the numbers of both DNA microsatellite alleles and mtDNA haplotypes were reduced in each of the four different selected lines (i.e the three mass selection lines and the winter mortality disease resistance line)

The diversity loss seemed to be greatest for one of the lines selected for WM resistance (B), a line that suffered the greatest selection pressure due to high annual losses to winter mortality Previously Nguyen (2009) also reported loss of microsatellite allele variation in other

S glomerata hatchery lines, different from those considered here (specifically in 4 th generation

S glomerata QX resistant family lines) Such losses were evident even though Nguyen (2009) used, for the most part, different DNA microsatellite loci from those considered here, and also he used a different wild reference line, although the sample sizes were equivalent in both studies In our study, the mtDNA analysis using COX1 confirmed the trend for a loss of genetic variation among all the selected lines

The loss in genetic diversity in long term mass selection or hatchery lines has been reported for many species, including Eastern oyster, Crassostrea virginica (Carlsson et al., 2006,

Yu and Guo, 2004) A 22-44% loss of alleles was reported for hatchery bred cultured silver-lipped pearl oysters, Pinctada maxima (Lind et al., 2009) A substantial loss of allele number in F1 progeny of hatchery bred abalone was reported for Haliotis rubra in Australia and Haliotis midae in South Africa (Evans et al., 2004a) and in 4 th generation selected C gigas lines (Appleyard and Ward, 2006)

Most large panmictic populations possess high levels of genetic diversity (Frankham,

1996), a group which we assume includes wild oysters However, the level of genetic diversity can be related to population size (Amos and Harwood, 1998) Our estimates of population size for the hatchery lines were relatively low, between 20-50 individuals The low Ne may have resulted from inadvertently selecting the best offspring produced by few parents or due to asymmetric reproduction and survival rates of broodstock (Lallias et al., 2010) Loss of variation can occur even in absence of directional breeding since effective population sizes can be reduced simply by drift resulting from the high variation of reproductive success known to exist in hatchery-propagated bivalves and shrimp (Boudry et al., 2002, Goyard et al., 2003)

Our estimates of population size generally conform to reports after mass selection in other aquaculture species including C gigas, where Ne was 9-36 (Appleyard and Ward, 2006), white leg shrimp Penaeus vannamei (Ne was 5-25 after 14 generations and 44-49 after 12 generations) (De Donato et al., 2005) So the various estimates of Ne here, with those reported in other species, tend to be lower than a minimum Ne of 50 recommended to limit inbreeding to around 1% per generation in a selection program (Bijma et al., 2000, Ponzoni et al., 2010)

A 1% level of inbreeding per generation is thought to be sustainable for breeding programs in terms of maintaining long term genetic response (Frankel and Soule, 1981) Our estimates of inbreeding (for a given generation), were an order of magnitude higher than those acceptable as sustainable, that is, the present hatchery, breeding and selection practices for NSW

S glomerata, at least those for mass selection, are not genetically sustainable High levels of

73 inbreeding were also found in other aquaculture species subject to mass selection, varying from 3-5% per generation, depending on species, number of generations and genetic management practices (Blonk et al., 2009, Brown et al., 2005)

There are a variety of consequences from inbreeding, and they can be considered in terms of a) impact on future selection response; and b) impact on the actual production animals Considering the former (point a), long term selection response can be diminished with inbreeding and inbreeding depression (Taris et al., 2007) For the latter (point b), Evans et al (2004b) estimated that a 10% increase in breeding coefficient (F) would cause 8.8% decrease in average body weight and 4.26% fall in survival in C gigas Therefore, loss of genetic variation may limit selection response for mass selected S glomerata in the future, but also impact on current commercial productivity Decrease in selection response for total weight after generations of mass selection was reported in S glomerata (Nell et al., 1999) and other species e.g European oyster, Ostrea edulis (Newkirk and Haley, 1983) and cold water fish (Gjedrem, 2000) Many other breeding programs failed to obtain genetic gains with mass selection e.g tilapia (Hulata et al.,

1986, Teichert-Coddington and Smitherman, 1988) and common carp, Cyprinus carpio (Moav and Wohlfarth, 1976)

What the prospects to maintain genetic variation for the long term in mass selected lines are? Recently, Knibb et al (2014b) suggested that if a line is split into various sublines, which are maintained separately, then while there is loss of variation in each line, the overall genetic among lines is still maintained , which would not be kept by maintaining one single large line This proposition at first seems to contradict genetic theory that since inbreeding is proportional to population size, one large population should become inbred more slowly than the various smaller sublines This may be correct, but only up until lines become or start to become fixed for alleles, at which time the sublines will maintain more variation, because each subline will tend to be fixed for different alleles Of course, if very few sublines are kept, then so too will limited variation be maintained

The process of selection, if intense, can accelerate fixation even in large populations Also, due the enormous fecundity of some aquacultured species, one or few families can be sampled by chance as the next broodstock which may lead to few families in the single line, but not in the different sublines, where chance sampling should choose different families Unequal family contributions have been noted in the Japanese flounder Paralichthys olivaceus where 99% of the offspring was found to be from just one single male while another two cohort males contributed

74 less than 1% (Sekino et al., 2004) Similarly, in the cultured silver-lipped pearl oyster, up to 40% of the progeny came from just a single family in spite of a large cohort of 28 parents used (Lind et al., 2009) In Ostrea edulis, mass spawning resulted in only a limited number of parents contributing to progeny, leading to the high inbred population comprising mostly the same half or fullsib offspring (Launey et al., 2001).

So while this postulate of maintaining genetic diversity by subdivision of lines was supported empirically with data from banana shrimp (Knibb et al., 2014), and supported by simulations (Lacy, 1987), it is not yet confirmed by independent observations in other species or cases In our present study for S glomerata there seems to be the right mixture of attributes to test Knibb’s proposition, namely a species that is highly fecund (single female S glomerata can produce millions of offspring), mass selection, intense at times, with no pedigree management, and last, an ancestral line that was split into various sublines that have been maintained for several generations separately without intercrossing Indeed, the pattern of allelic and haplotype variation within and among the S glomerata lines are not inconsistent with those predicted by Knibb’s proposition, namely, loss of alleles and haplotypes in individual lines, but retention of substantial variation among lines Considering all four S glomerata sublines together, the number of different alleles was not significantly different from the wild, and should more sublines have been kept, it is possible even more variation would also have been kept A caveat of this approach is that in some cases extra costs may be needed to maintain multiple lines

There are various other strategies discussed elsewhere how to maintain genetic diversity without pedigree information The simplest perhaps is to offset loss by reintroducing variation again from the wild, but there are disadvantages from doing this Wild animals are genetically unimproved, so their addition to the breeding nucleus could degrade genetic selection responses already achieved Moreover, the introduction of wild stock risks the introduction of pathogens Bentsen and Olesen (2002) have proposed a modified mass selection based on pair mating and the contribution of each mate to the next generation Nell (2006) suggested for S glomerata the use of several small spawnings with equal number of males and females in each cohort Evans et al (2004b) considered cross breeding between distantly related lines or even with the wild animals to avoid increased inbreeding in selective bred bivalve species Notwithstanding these efforts to preserve diversity without pedigrees, it is most likely that maximum diversity can only ever be maintained in the very long term by using full pedigree information (Evans et al., 2004b, Knibb et al., 2014b) One variation of pedigree based selection is called “walk back” selection,

Conclusion

Seven microsatellite DNA markers were newly developed and qualified for analysis of S glomerata genetic diversity There was a substantial loss of allele and haplotype diversity of each of the four selected lines of S glomerata relative to the wild samples Pooling across hatchery lines, the level of allelic and haplotype diversity was not dissimilar to that of the wild The present data support the proposal of Knibb et al (2014) that genetic diversity in a mass selection program can be maintained by setting up a number of different independent sublines.

Acknowledgements

We gratefully acknowledge financial support from the University of the Sunshine Coast (USC) and Australian Centre for International Agricultural Research (ACIAR) through the John Allwright Fellowship to V-V I We also acknowledge the Port Stephens Fisheries Institute, NSW and its staff for providing samples and their assistance in sampling We highly appreciate the technical assistance from USC staff and students, especially David Bright, Rob Lamont, Ido Bar and Dan Powell in instruction of lab techniques and genetic software.

Supplementary data

Supplementary file S2 Allelic and haplotypic data of Sydney rock oyster, Saccostrea glomerata

References

Amos, W., Harwood, J., 1998 Factors affecting levels of genetic diversity in natural populations

Amos, W., Balmford, A., 2001 When does conservation genetics matter? Heredity 87, 257-265

Appleyard, S.A., Ward, R.D., 2006 Genetic diversity and effective population size in mass selection lines of Pacific oyster (Crassostrea gigas) Aquaculture 254, 148-159

Astanei, I., Gosling, E., Wilson, J., Powell, E., 2005 Genetic variability and phylogeography of the invasive zebra mussel, Dreissena polymorpha (Pallas) Molecular Ecology 14, 1655-1666

Banks, S.C., Piggott, M.P., Raftos, D.A., Beheregaray, L.B., 2006 Microsatellite markers for the Sydney rock oyster, Saccostrea glomerata, a commercially important bivalve in South Eastern Australia Molecular Ecology Notes 6, 856-858

Bentsen, H.B., Olesen, I., 2002 Designing aquaculture mass selection programs to avoid high inbreeding rates Aquaculture 204, 349-359

Bijma, P., Van Arendonk, J.A., Woolliams, J.A., 2000 A general procedure for predicting rates of inbreeding in populations undergoing mass selection Genetics 154, 1865-1877

Blonk, R.J., Komen, J., Kamstra, A., Crooijmans, R.P., van Arendonk, J.A., 2009 Levels of inbreeding in group mating captive broodstock populations of Common sole, Solea, inferred from parental relatedness and contribution Aquaculture 289, 26-31

Boudry, P., 2008 Review on breeding and reproduction of Europhean aquaculture species Aqua

Boudry, P., Collet, B., Cornette, F., Hervouet, V., Bonhomme, F., 2002 High variance in reproductive success of the Pacific oyster (Crassostrea gigas, Thunberg) revealed by microsatellite-based parentage analysis of multifactorial crosses Aquaculture 204, 283-296

Brown, R.C., Woolliams, J.A., McAndrew, B.J., 2005 Factors influencing effective population size in commercial populations of gilthead seabream, Sparus aurata Aquaculture 247, 219-225

Carlsson, J., Morrison, C.L., Reece, K.S., 2006 Wild and Aquaculture Populations of the Eastern Oyster

Compared Using Microsatellites Journal of Heredity 97, 595-598

De Donato, M., Manrique, R., Ramirez, R., Mayer, L., Howell, C., 2005 Mass selection and inbreeding effects on a cultivated strain of Penaeus (Litopenaeus) vannamei in Venezuela Aquaculture 247, 159-167

Dove, M.C., O'Connor, W.A., 2009 Commercial assessment of growth and mortality of fifth generation

Sydney Rock Oysters, Saccostrea glomerata (Gould, 1850) selectively bred for faster growth Aquaculture Research 40, 1439-1450

Evans, B., Bartlett, J., Sweijd, N., Cook, P., Elliott, N., 2004a Loss of genetic variation at microsatellite loci in hatchery produced abalone in Australia, Haliotis rubra and South Africa, Haliotis midae Aquaculture 233, 109-127

Evans, F., Matson, S., Brake, J., Langdon, C., 2004b The effects of inbreeding on performance traits of adult Pacific oysters, Crassostrea gigas Aquaculture 230, 89-98

Folmer, O., Black, W.H., Hoeh, W., Lutz, R., Vrijenhoek, R., 1994 DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates Molecular Marine Biology and Biotechnology 3, 294-299

Frankel, O.H., Soule, M.E., 1981 Conservation and Evolution Cambridge University Press, Cambridge, UK

Frankham, R., 1996 Relationship of genetic variation to population size in wildlife Conservation Biology

Gene Codes Corporation, I., 2011 Sequencher version 5.0 sequence analysis software Gene Codes

Gjedrem, T., 2000 Genetic improvement of cold-water fish species Aquaculture research 31, 25-33 Goyard, E., Arnaud, S., Vonau, V., Bishoff, V., Mouchel, O., Pham, D., Wyban, J., Boudry, P., 2003 Residual genetic variability in domesticated populations of the Pacific blue shrimp (Litopenaeus stylirostris) of New Caledonia, French Polynesia and Hawaii and some management recommendations Aquatic Living Resources 16, 501-508

Hulata, G., Wohlfarth, G.W., Halevy, A., 1986 Mass selection for growth rate in the Nile tilapia,

Jones, O.R., Wang, J., 2010 COLONY: a program for parentage and sibship inference from multilocus genotype data Molecular Ecology Resources 10, 551-555

Kincaid, H.L., 1983 Inbreeding in fish populations used for aquaculture Aquaculture 33, 215-227

Knibb, W., Whatmore, P., Lamont, R., Quinn, J., Powell, D., Elizur, A., Anderson, T., Remilton, C., Nguyen,

N.H., 2014 Can genetic diversity be maintained in long term mass selected populations without

77 pedigree information? — A case study using banana shrimp Fenneropenaeus merguiensis

Knibb, W., Miller, A., Quinn, J., D’Antignana, T., Nguyen, N.H., 2016 Comparison of lines shows selection response in kingfish (Seriola lalandi) Aquaculture 452, 318–325 doi:10.1016/j.aquaculture.2015.11.015

Lacy, R.C., 1987 Loss of genetic diversity from managed populations: interacting effects of drift, mutation, immigration, selection, and population subdivision Conservation Biology 1, 143-158

Lallias, D., Boudry, P., Lapegue, S., King, J.W., Beaumont, A.R., 2010 Strategies for the retention of high genetic variability in European flat oyster (Ostrea edulis) restoration programmes Conservation Genetics 11, 1899-1910

Launey, S., Barre, M., Gerard, A., Naciri-Graven, Y., 2001 Population bottleneck and effective size in

Bonamia ostreae-resistant populations of Ostrea edulis as inferred by microsatellite markers Genetical research 78, 259-270

Li, G., Hubert, S., Bucklin, K., Ribes, V., Hedgecock, D., 2003 Characterization of 79 microsatellite DNA markers in the Pacific oyster Crassostrea gigas Molecular Ecology Notes 3, 228-232

Lind, C.E., Evans, B.S., Knauer, J., Taylor, J.J., Jerry, D.R., 2009 Decreased genetic diversity and a reduced effective population size in cultured silver-lipped pearl oysters, Pinctada maxima Aquaculture

Lopera-Barrero, N.M., Povh, J.A., Ribeiro, R.P., Gomes, P.C., Jacometo, C.B., da Silva Lopes, T., 2008

Comparison of DNA extraction protocols of fish fin and larvae samples: modified salt (NaCl) extraction Ciencia e Investigación Agraria 35, 65-74

Marshall, T., Slate, J., Kruuk, L., Pemberton, J., 1998 Statistical confidence for likelihood-based paternity inference in natural populations Molecular ecology 7, 639-655

Moav, R., Wohlfarth, G., 1976 Two-way selection for growth rate in the common carp, Cyprinus carpio

Nell, J.A., 2006 Manual for mass selection of Sydney rock oysters for fast growth and disease resistance

Fisheries Research Report series 13 NSW Department of Primary Industries, NSW, pp 53

Nell, J.A., Perkins, B., 2005 Evaluation of progeny of fourth generation Sydney rock oyster, Saccostrea glomerata (Gould, 1850) breeding lines Aquaculture Research 36, 753-757

Nell, J.A., Smith, I.R., Sheridan, A., 1999 Third generation evaluation of Sydney rock oyster Saccostrea commercialis (Iredale and Roughley) breeding lines Aquaculture 170, 195-203

Newkirk, G., Haley, L., 1983 Selection for growth rate in the European oyster, Ostrea edulis: response of second generation groups Aquaculture 33, 149-155

Nguyen, D., 2009 Assessing genetic diversity in cultured aquatic species: the Sydney Rock Oyster,

Saccostrea glomerata stock improvement program as a model, Faculty of Science and

Technology Queensland University of Technology

O'Connor, W.A., Michael, D., Ben, F., S., O.C., 2008 Manual for hatchery production of sydney rock oysters, Saccostrea glomerata IDP NSW NSW Department of Primary Industries Port Stephens Fisheries Center Taylors Beach NSW 2316, Australia http://www.dpi.nsw.gov.au/research/areas/aquaculture/outputs/2008/oconnor3

Peakall, R., Smouse, P.E., 2012 GenAlEx 6.5: genetic analysis in Excel Population genetic software for teaching and research-an update Bioinformatics 28, 2537-2539

Ponzoni, R.W., Khaw, H.L., Nguyen, N.H., Hamzah, A., 2010 Inbreeding and effective population size in the Malaysian nucleus of the GIFT strain of Nile tilapia, Oreochromis niloticus Aquaculture 302, 42-48

Pritchard, J.K., Wen, W., Falush, D., 2003 Documentation for STRUCTURE software: version 2 http://pritchardlab.stanford.edu/software/readme_structure2.pdf

Reece, K., Ribeiro, W., Gaffney, P., Carnegie, R., Allen, S., 2004 Microsatellite marker development and analysis in the eastern oyster (Crassostrea virginica): confirmation of null alleles and non-

Mendelian segregation ratios Journal of Heredity 95, 346-352

Schmidt, R., 2011 Evaluation of GeneMapper® ID-X and GeneMarker® HID for use at the NYC OCME

Sekino, M., Sugaya, T., Hara, M., Taniguchi, N., 2004 Relatedness inferred from microsatellite genotypes as a tool for broodstock management of Japanese flounder, Paralichthys olivaceus Aquaculture

Sonesson, A.K., 2005 A combination of walk-back and optimum contribution selection in fish: a simulation study Genetics Selection Evolution 37, 587–599 doi:10.1051/gse:2005020

Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kumar, S., 2011 MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods Molecular Biology and Evolution 28, 2731-2739

Taris, N., Batista, F.M., Boudry, P., 2007 Evidence of response to unintentional selection for faster development and inbreeding depression in Crassostrea gigas larvae Aquaculture 272, S69-S79 Taris, N., Ernande, B., McCombie, H., Boudry, P., 2006 Phenotypic and genetic consequences of size selection at the larval stage in the Pacific oyster, Crassostrea gigas Journal of Experimental Marine Biology and Ecology 333, 147 - 158

Teichert-Coddington, D.R., Smitherman, R.O., 1988 Lack of response by Tilapia nilotica to mass selection for rapid early growth Transactions of the American Fisheries Society 117, 297-300

Trenaman, R., Livingstone, S., Creese, A., 2014 Aquaculture Production Report, NSW DPI, 10pp

Van Oosterhout, C., Hutchinson, W.F., Wills, D.P., Shipley, P., 2004 MICRO-CHECKER: software for identifying and correcting genotyping errors in microsatellite data Molecular Ecology Notes 4, 535-538

Yu, Z., Guo, X., 2004 Genetic Analysis of Selected Strains of Eastern Oyster (Crassostrea virginica Gmelin)

Using AFLP and Microsatellite Markers Marine Biotechnology 6, 575-586

Chapter 4: Is the oyster cultivated in Vietnam the Pacific oyster Crassostrea gigas and can current aquaculture practices maintain genetic variation among lines? A case study on taxonomy and genetic diversity for the establishment of a breeding program

Vu Van In a,b , Wayne O’Connor, Vu Van Sang b , Phan Thi Van c and Wayne Knibb a,e

Published on Aquaculture a University of the Sunshine Coast, Maroochydore, Queensland 4558, Australia b Northern National Broodstock Center for Mariculture, RIA1, Cat Ba islands, Haiphong, Vietnam c Research institute for Aquaculture no1 (RIA1), Tuson, Bacninh, Vietnam d Industry and Investment NSW, Department of Primary Industry, Port Stephens Fisheries Institute, Taylors Beach, NSW, 2316, Australia e Corresponding author at: Genecology Research Center, Faculty of Science Health and Education, University of the Sunshine Coast, Locked Bag 4, Maroochydore Dc, QLD 4558, Australia Tel.:+61 7 5430

2831 E-mail address: wknibb@usc.edu.au, vuvanin@ria1.org

Oyster aquaculture is an emerging and rapidly growing aquaculture sector in northern Vietnam, but there is confusion about the identity of species presently under culture Several stock introductions have occurred and there is conjecture as to whether current cultured populations are Pacific oysters Crassostrea gigas or the Portuguese oyster Crassostrea angulata Here we aim to resolve this confusion and assess whether, after generations of captive breeding, there has been substantial loss of genetic variation that could impact long-term selection response via selective breeding programs

To address the first aim of this study we searched for nucleotide differences in the mitochondrial DNA cytochrome c oxidase subunit 1 (COX1) that, for the first time, would categorically separate and distinguish C angulata and C gigas using large number of haplotype references (288: 222 for C gigas vs 56 for C angulata) from GenBank Specifically, we reviewed

300 published haplotypes of C angulata and C gigas based on a 293 bp nucleotide-fragment of published COX1 sequences, and found that there were five distinct nucleotides that are categorically different between C angulata and C gigas that can be considered as diagnostic

Is the Vietnam aquaculture Pacific oyster Crassostrea gigas? and can current

Introduction

Several attempts have been made to establish an oyster industry in Vietnam Over a decade ago Pacific oysters, Crassostrea gigas, were introduced to central Vietnam from Australia (Heasman et al., 2000), but after initial hatchery success, stocks died before harvest In 2007, another attempt to establish an industry was made, this time in northern Vietnam using existing oyster stocks, reported to be C gigas This latter attempt proved successful with production growing rapidly to over 7,000 tonnes/annum in 2010 with the majority coming from Quangninh and Haiphong provinces However, despite the industry growth, considerable doubt existed among researchers regarding the phylogenetic status of the oyster being cultivated In part the doubt arose from the fact that the original broodstock were imported from Taiwan, where C angulata, is thought to underpin production (Batista et al., 2005, Boudry et al., 1998) Since 2007, spat from southern China have been imported to satisfy the increasing demands for oyster production These spat may consist of C gigas as China is a major producer of C gigas (Boudry

81 et al., 2003, Lapegue et al., 2004), although mixed populations of C angulata and C gigas in northern China have been reported (Batista et al., 2005) Therefore, it was unclear whether the Vietnam’s hatchery oyster stocks were C angulata or C gigas, a hybrid or possibly another species

C gigas and C angulata cannot be accurately distinguished by morphological appearance In the past, C angulata and C gigas were first considered to be different species on the basis of geographical distribution, while C angulata was first found along the coasts of southern Europe, C gigas was naturally distributed in Asia (Soletchnik et al 2002) Later, they were considered to be a single species due to their phenotypic similarity, fertile hybrid offspring (Huvet et al., 2004a, Huvet et al., 2002)

To distinguish between oyster species, percent differences in mtDNA sequences, especially COX1, have been widely used as they are conserved within species but vary even among close taxa (Boudry et al., 1998, Foighil et al., 1998, Huvet et al., 2000a) Examples of their use include differentiating between the flat oyster Ostrea edulis and the native Ostrea angasi in Western Australia (Morton et al., 2003), C iredalei and Saccostrea cucullata in Thailand

(Klinbunga et al., 2003), and Crassostrea oyster species: C iredalei, C belcheri and C madrasensis in Malaysia (Mustaffa et al., 2010) COX1 has also been used to discover new oyster species in China e.g Crassostrea hongkongensis (Lam and Morton, 2003) Several authors have suggested

C gigas and C angulata could be separated using percent differences in COX1 sequences

(Boudry et al., 1998, Huvet et al., 2000a, Huvet et al., 2000b, Wu and Yu, 2009), even though the differences are only a few percent, but others (Boudry et al., 2003, Liu et al., 2011) point out that such slight differences are lower than those that normally evident between species, so the taxonomic status of these two species remains controversial

Other mtDNA sequences (David and Savini, 2011), nuclear nucleotide sequences (Menzel,

1974), DNA microsatellite alleles (Reece et al., 2008) and allozymes (Boudry et al., 1998) did not distinguish between C gigas and C angulata Little nucleotide difference in nuclear rRNA ITS1 regions and small mitochondrial phylogenetic divergence has been observed (Menzel, 1974) Studies using allozyme markers (Boudry et al., 1998) or microsatellite DNA markers (Huvet et al., 2000a, Reece et al., 2008) showed no significant differences between C giga and C angulata

Other mtDNA genes: rrnL and MNR, 16S rDNA, 12S rDNA (Boudry et al., 2003, David and Savini, 2011, Lam and Morton, 2006, Masaoka and Kobayashi, 2005, Stepien et al., 2001) and

82 nuclear genes: ITS1, ITS2 18S and 28S rDNA (David and Savini, 2011, Larsen et al., 2005) have been used to analyse the phylogeny of bivalves but they are less diverse than COX1 and even less effective in separating C gigas and C angulata (Boudry et al., 2003, Radulovici et al., 2010,

Stepien et al., 2001) Although the Internal transcribed spacer (ITS) region of mtDNA has a high degree of variation among some species of Ostreidae, it too cannot separate C gigas and C angulata (Wang and Guo, 2008)

Beyond the taxonomic uncertainty of the oysters being cultured in Vietnam, there are additional concerns that with captive reproduction of such fecund species, that lines could go through population bottlenecks and lose genetic diversity, yielding inbred and poor quality spat

To develop and sustain an oyster industry in the long term, and to employ in selective breeding, it is desirable to use outbred stocks and to sustain their genetic diversity in the future Maintaining a wide range of genotypes could give a hatchery population more flexibility of response to a constantly changing environment (Boudry, 2008, Taris et al., 2006) Genetic diversity is the initial requirement for a genetic improvement program, however it may be eroded by the process of selection and husbandry practices due to a limited number of broodstock individuals and high variation in individual reproductive success (Boudry, 2008, Nguyen, 2009, Taris et al., 2006) With a high rate of inbreeding, individuals mate with their close relatives which results in a higher incidence of recessive deleterious genotypes Previous studies reported loss of genetic variation in many hatchery populations, especially for broadcast spawners like oysters where a female oyster can release millions of egg in only one spawning event (O'Connor et al., 2008b) The Vietnamese hatchery oyster stocks have been established and bred for almost seven generations (seven years from 2008-2014), and there is concern about loss of variation and inbreeding

Here we investigate three main issues 1) can DNA sequence divergence data separate C gigas and C angulata or do we need other analyses, 2) are the oysters being farmed in Vietnam

C gigas? And 3) are the current hatchery stocks sufficiently genetically diverse and adequate to form the basis of a selective breeding program?

Materials and methods

The four oyster stocks used in this study were collected from three grow-out regions of Vietnam: Vandon, Catba and Nhatrang (Figure 4.1) The stocks included 1) RIA1, a research population cultured in Vandon, 2) China, a stock imported from southern China cultured in Vandon, 3) Namdinh, the stock cultured at Catba produced by local Namdinh hatcheries, and 4) Nhatrang, local oysters farmed at Nhatrang in central Vietnam

All samples from the four populations were taken in December

2013, and they were preserved in 70% ethanol and shipped to University of the Sunshine Coast

(USC), where they were stored at -20 o C until required The minimum sample size for analysis of

DNA microsatellite loci and COX1 was 32 and 20 individuals for each population, respectively

The animals were randomly picked up for labelling by numbering, and so truncating the list to a common number for all populations serves functionally as randomly sampling

Figure 4.1 Hatchery and sampling sites

DNA extraction was according to the NaCl extraction protocol (Lopera-Barrero et al.,

2008) The integrity of the DNA was verified by horizontal electrophoresis in a 0.9% agarose gel, at 110 volts for 40 minutes a 0.6xTBE buffer (500 mM Tris-HCl, 60 mM boric acid, and 83 mM EDTA) The gel was stained with ethidium bromide, verified and captured in GeneSnap with the Syngene System Bio-Rad) Moreover, the quality and quantity of the DNA were then evaluated using a Nanodrop 2000 (Thermo Scientific, USA) at the absorbance of 260/280nm Good DNA templates were then diluted in molecular grade water (Amresco) to 25 ng/àL -1

4.2.4 Mitochondrial DNA sequencing and analysis

The mitochondrial cytochrome C oxidase subunit I (COX1) was amplified using universal primers (LCO1490 and HCO2198) developed by Folmer et al (1994) in 25 àL reactions using MyTaq DNA polymerase (Bioline) PCR reactions containing 2xMyTaq reaction buffer, 0.25 mg ml -1 of bovine serum albumin, 60 ng of template DNA, 0.04 àM of forward and reverse primers where forward primers labelled with fluorescent dyes (FAM, NED, PET or VIC), and 2 unit of MyTaq DNA polymerase (Bioline) Ninety six DNA samples from four different cultured lines (19-

20 samples per each population) were used for COX1 amplification and sequencing using an ABI 3730XL DNA analyser Sequencher 5.0 (Gene Codes Corporation, 2011) was used to correct errors in COX1 sequences and trim them Mega 6.06 (Tamura et al., 2013), was used for sequence alignment, analysis of nucleotide differences, identification of haplotypes and building phylogenetic tree

All available published COX1 sequences (published references) of C gigas and C angulata from GenBank were downloaded as Fasta files and analysed to determine whether there were any fixed nucleotide differences among these two species for their COX1 sequences using Mega 6.06 (Tamura et al., 2013) and GeneDoc 2.7 (Nicholas et al., 1997) A phylogenetic tree was built using all of the sample COX1 sequences and some published references of C gigas and C angulata in Mega 6.06 (Tamura et al., 2013)

To identify the oysters sampled from Nhatrang, published COX1 sequences of other

Crassostreinae (namely, C sikamea, C ariakensis, C hongkongensis, C brasiliana, C virginica, C belcheri, C nippona, C iredalei and C madrasensis) and the Ostreidae (O edulis, O chilensis and

O aupouria) from GenBank were added and aligned using COX1 references using in Mega 6.06

(Tamura et al., 2013), and then a tentative phylogenetic tree was built to investigate what species the study samples clustered with For a final phylogenetic tree, only published COX1 references were used together with COX1 sequences from samples in this study for constructing a neighbour join tree on Mega 6.06 (Tamura et al., 2013)

From the published microsatellite primers, 48 primer pairs were selected (based on the polymorphic information content, number of alleles, expected heterozygosity), and used to amplify the farmed oyster individuals in Vietnam in 12.5àL PCR reactions containing 1xMyTaq reaction buffer, 0.125 mg ml -1 of bovine serum albumin, 30 ng of template DNA, 0.02 àM of forward and reverse primers where forward primers labelled with fluorescent dyes (FAM, NED, PET or VIC), and 1 unit of MyTaq DNA polymerase (Bioline) The reaction was amplified in an Eppendorf Mastercycler gradient thermal cycler using the following steps: an initial denaturing for 5 min at 95 o C followed by 30 cycles of 95 o C for 30 secs, annealing for 15s, extension at 72 o C for 10 secs and final extension at 72 o C for 2 mins PCR products were qualified on 3% agarose gels run for 2 h at 140 volts to detect levels of polymorphism (high variation of alleles on a locus) and quality of amplified products On the basis of potential polymorphisms evident from agarose gel electrophoresis, 12 putatively polymorphic microsatellites were tested on eight different individual samples using fluorescent labelled primers and genotyped in an AB 3500 Genetic Analyser, Hitachi Nine out of 12 published primers were qualified for this study based on good amplification, repeatability and reliability of scoring and degree of allelic variation (Supplementary file S3_Table 2)

PCR products were genotyped using an AB 3500 Genetic Analyser For preparation prior to genotyping, PCR products were diluted 10 to 50 fold depending on the yield of PCR amplified products One àL of each diluted PCR product was then transferred to 9 àL of a combination of Liz 600 (size standard V2.0) and HiDi formamide from Life Technologies (7 àL Liz for 500 àL HiDi), then vortexed and briefly centrifuged before denaturation for 5 min at 95 o C, cooling of DNA samples for 2-3 mins in an ice bath in order to fix the DNA before transferring them to the genotyping machine Genotyping was conducted using the 500bp fragment analysis program

4.2.7 Data analysis and statistical methods

Output (.fsa files) from the genotyper were analysed and scored using GeneMarker 2.6.3 (Schmidt, 2011) and then scoring was confirmed manually Genotyping data was checked for null allele on Microchecker 2.2.0.3 (Van Oosterhout et al., 2004) Genalex 6.5 (Peakall and Smouse,

2012) was used to estimate Fst and analyse allele diversity, molecular variance, fixation index (Fis) Cervus 3.0 (Marshall et al., 1998) was used to run Hardy-Weinberg equilibrium test and estimate polymorphic information content (PIC) and null allele frequency Colony 2.0.2.3 (Jones and Wang, 2010) was used to determine sample size, inbreeding and parentage assignment SPSS statistics 22 was used to conduct Chi-square tests (SPSS statistics 22, 2013)

Genetic relatedness among individuals from five groups was estimated by principal coordinates analysis, program setting for K = 1, 10 iterations on Structure 2.2 (Pritchard et al.,

Results

4.3.1 ID identity and mtDNA haplotype diversity among lines

Part of the analysis on 288 published COX1 sequences consisting of 222 sequences of C gigas, 56 sequences of C angulata obtained from GenBank (full data in supplementary file

S3_Table 1) is shown in Table 4.1 There are five fixed nucleotide differences between C angulata and C gigas COX1 sequences on a 293 nucleotide residue (namely, at position 88 -T vs C, position

105 -C vs T, position 138 -T vs C, position 193 -T vs C and position 264 - G vs A) Interestingly, the well-known C angulata haplotypes A (GenBank assess no: AJ553907.1), B (AJ553908.1) used as COX references carry all these five exclusive nucleotides for C angulata, meanwhile the most common C gigas haplotypes C (AJ553909.1) and E (AJ553911.1) used as references contains all five exclusive nucleotides for C gigas (Table 4.2)

For identification of oyster ID in northern Vietnam, sequence alignment with the above COX1 references with 72 samples was analysed, showing all of them from three hatchery lines (Ria1, Namdinh and China) carry all of the five diagnostic nucleotides indicative of C angulata (Table 4.2), of those three new C angulata haplotypes were found, consisting of haplotype i1 (6 samples in Namdinh), haplotype i2 (3 samples in Namdinh and 6 samples in China line) and haplotype i3 (one sample in China line) (Table 4.2)

The phylogenetic tree shows COX1 references including the most common haplotype A of C angulata definitely separating from C gigas including the most common haplotype C, whereas COX1 of all samples from three lines (Ria1, Namdinh and China) in northern Vietnam belongs to C angulata

Table 4.1 Five exclusive nucleotides in COX1 sequences between C gigas and C angulata Gray shading indicates the exclusive nucleotide

Table 4.2 Alignment of oyster sampled from lines of Ria1, Namdinh and China with COX1 references of C gigas and C angulata *: new C angulata haplotype i1; **: new C angulata haplotype i2;

***: new C angulata haplotype i3 The numbers after the name of each line indicates sample numbers The different letters (a, b and c) after the sample numbers indicate the different haplotypes.

4.3.2 Identification of hatchery oyster ID in Nhatrang, Vietnam

Among the 16 samples from the Nhatrang line, there were four different types of haplotypes When we aligned them with counterparts of other species, they were obviously different from C angulata and C gigas Instead, there were two haplotypes (haplotype 1 and 2,

Table 4.3) corresponding to C sikamea (AB641328; HQ661017), and the other two haplotypes correspond to C madrasensis (FJ428750: 01 sample; and JF915457: 9 samples, Table 4.3) Figure

4.2 shows that the local hatchery oysters in Nhatrang were clearly different from C gigas nor C angulata, they were a mix of two species: C sikamea and C madrasensis

Table 4.3 Alignment of oyster sampled from Nhatrang with Crassostrea references Numbers after the name of each line indicates sample numbers, the different numbers (1, 2, 3, 4) in the bracket, after the sample numbers, indicates the different haplotypes

Figure 4.2 Phylogenetic tree using MtDNA COX1 using phylogenetic reconstruction analysis with Neighbour-joining statistical methodusing Bootstrap test with 500 replications, and mode with maximum composite likelihood on Mega6.0 Number on the branches indicate bootstrap values Numbers after the name of the lines inside in brackets indicate samples (labelled by numbering), * indicates new C angulata haplotypes identified in this study

AB748801.1|_C_angulata A-Cr01 Nam_Dinh_(3_7_10_12_13_15)*

AB748805.1|_C_angulata_G-Cr03 AJ553907.1|_C_angulata_hap_A Ria1_(2_4_5_16_20_21_22_23_24) Namdinh_(1_2_8_11_17_18_20) China_(8_15)

AF152567.1|_C_angulata AB748802.1|_C_angulata_A-Cr05 Ria1_(1_3_6_7_10_12_14_15_18) Namdinh_(9_19_23_24)

AJ553911.1|_C_gigas_hap_E gAJ553909.1|_C_gigas_hap_C KJ801546.1|_C_gigas_hap_AS1 AB748797.1|_54_C_gigas_A-Cr06

HQ661017.1|_C_sikamea Nhatrang_(10_13_15_18) AB641328.1|_C_sikamea Nhatrang_(14_17) FJ428750.1|_C_madrasensis Nhatrang_8

4.3.3 Genetic diversity among C angulata lines

There are six haplotypes in 72 samples from hatchery lines (Ria1, Namdinh and China), of which three are new haplotypes (haplotype i1, 2 and 3) that are different from all published haplotypes in GenBank (Table 4.4) Namdinh and China line both have five haplotypes, including two out of three new haplotypes while RIA1 has the least haplotypes (3 haplotypes) and none are new haplotypes Interestingly, the number of haplotypes increased to six when all samples are pooled together (Pooled 72), or the pooled samples were adjusted down to 24 (Pooled 24) to be the same sample size as those of each line (Table 4.4) However, there is no significant difference between the number of different haplotypes in each of the hatchery line with either Pooled 72 or Pooled 24 sample (P>0.05)

Table 4.4 Number of mtDNA haplotypes in hatchery lines

Data presents the numbers of samples that carry the same haplotype referred to the column “Haplotype”, grey shading indicates private haplotypes

† : using 24 samples from Ria1, Namdinh and China

‡ : using all samples from Ria1, Namdinh and China

( a ) The same letters on the same row indicate no significant difference (P>0.05)

Using 9 loci to analyse genetic diversity of three hatchery lines, we found that the number of alleles ranged from 4 to 18 alleles per locus (Table 4.5) considering each hatchery separately (sample size of 32 for each line) and all samples were pooled together (sample size of 96, termed Pooled 96), and the pooled samples were adjusted down to 32 (Pooled 32) to be the same sample size as those of each line (this sample size was needed to avoid bias, i.e a greater likelihood of more alleles when there are more samples) (Table 4.5) It is evident that the pooled samples tend to have a greater number of alleles at almost every locus compared with each individual hatchery line, however there is not significant difference between each hatchery line and the pooled samples, regardless of pooling the 96 samples or using the pooled subset of 32 samples (P>0.05, Table 4.5 and Figure 4.3)

Table 4.5 Number of DNA microsatellite alleles among loci from Ria1, Namdinh and China lines

Pooled 32 † Pooled 96 ‡ Ria1 Namdinh China

Lo ci ( nu m be r o f alle le s)

Total number of different alleles a 83 a 88 a 90 a 98 a 121 a

Average No different ealleles per locus

† : using 32 samples in total from Ria1, Namdinh and China

‡ : using all samples from Ria1, Namdinh and China

( a ) the same letters on the same row indicate no significant difference (P>0.05)

Figure 4.3 Total number of alleles across nine loci in three cultured lines Different letters on top of the same type of bars represent no significant differences using Chi-square tests of counts (P>0.05) 1 : Number of alleles using 32 samples; 2 : Number of haplotypes using 24 samples; 3 : Number of alleles using 72 samples; 4 : Number of haplotypes using 96 samples

4.3.4 Inbreeding and effective population size assessed by analysis of genotypes using Colony software 2.0.2.3

Full sib families (i.e individuals that shared the same sire and dam) were detected in all three lines However, by running all the genotypes from three lines together, we found no evidence of full sibs shared across lines Accordingly, there is no evidence of cross contamination among different hatchery lines

Number of alleles Number of haplotypes

Table 4.6 Number of full-sibs and effective population size among lines

Total number of full-sib family

Effective population size with upper and lower 95% confidence intervals (N e )

† : using 32 samples in total from Ria1, Namdinh and China

†† : using all samples from Ria1, Namdinh and China

Genetic relationships among lines are shown in Table 4.7 and Figure 4.4 The genetic difference using Fst among three lines was small, greatest between Namdinh vs China (0.39), followed by Ria1 vs China (0.036) and lowest between Ria1 vs Namdinh (0.034) (Table 4.7) This is reiterated by Principal Coordinates Analysis for Genetic relationships among lines and individuals of the three hatchery lines, where samples of Ria1 and Namdinh distribute closer than those of China (Figure 4.4) There are some samples from lines clustered together, however the hatchery lines tended to cluster apart (Figure 4.4)

Table 4.7 Pairwise Population Fst value

Figure 4.4 Genetic distance among individuals of three lines from Principal Coordinates Analysis (PCA) using genetic distance matrices Individuals from the three populations are indicated by the colours and symbols shown “Axis 1 (26.64%)” means the genetic variation between samples located on the right and the left sides of the Axis 1 is 25.64%; “Axis 2 (18.99%)” is 18.99% genetic difference between samples distributed on the upper and lower sites of the horizontal Axis 2.

Discussion

Species specific fixed nucleotide differences for C angulata and C gigas

Previous attempts to resolve the taxonomy of the two closely related species, C angulata and C gigas using DNA microsatellite alleles and allozymes have been unsuccessful (Boudry et al., 1998, Huvet et al., 2000a, Reece et al., 2008), while the ability to separate them based on percentages of mtDNA COX1 haplotypes or DNA sequence divergence has been controversial

An issue seems to be the limited the divergences for these two taxa which is very slight e.g 2-3% (Boudry et al., 2003) and 2.22 – 3.37% (Wang et al., 2010), far lower than 13-14% divergence between C angulata and C hongkongensis or 12-13% divergence between C gigas and C nippona (Boudry et al., 2003) In addition, Liu et al (2011) reported 2.2-3.2% genetic distances between C gigas and C angulata while the variation of COX1 between species ranges from 9- 30% within oyster genus or 25-32.5% between species from different genera based on pairwise divergence Thus, the COX1 sequence divergence between C gigas and C angulata is lower than what is normally evident between oyster species (Liu et al., 2011, Wang et al., 2010) raising a question about the species status of C angulata and C gigas – are they one species with slightly divergent populations or two species?

O'Foighil et al (1998) were the first authors to attempt to differentiate the two taxa by looking for fixed nucleotide differences at particular positions They reported 11 fixed nucleotide differences between C angulata and C gigas, and concluded that they were two species, not a single species as reported by many previous reports However, the limitation of the work by O’Foighil et al 1998 is that a small number of haplotypes were analysed (because few were available at that time) preventing a firm conclusion about the existence of exclusive nucleotides between C angulata vs C gigas O’Foighil et al 1998 used one C gigas haplotype (AF152565) - identical to the most common known C gigas haplotype C (Huvet et al., 2000b) and four C angulata haplotypes: angul1, angul3, angul4 and angul2 (the only one was published on GenBank at the time: AF152567) Subsequently, Huvet et al., (2000b) used an analogous approach and suggested there could be specific mtDNA haplotypes unique to each species, such as haplotype

A for C angulata and haplotype C for C gigas However, the problem is that not all oyster individuals carry these haplotypes, rather they may carry one of various types (to date 85 different haplotypes for C gigas and 18 different haplotypes for C angulata have been reported (Huvet et al., 2000b)

This study builds on the historical work of O'Foighil et al (1998) and analyses 222 COX1 references of C gigas and 56 COX1 sequences of C angulata from GenBank to reveal there are

99 at least five completely exclusive nucleotides between the two taxa Four of the five exclusive nucleotides found in this study (namely, at position 88 -T vs C, position 105 -C vs T, position 138 -T vs C, and position 264 - G vs A on 293 bp residue) likely corresponds to four distinct nucleotides discovered by O'Foighil et al (1998) (namely, at position 368 –T vs C, position 384 –

C vs T, position 417 –T vs C, 559 –G vs A on 588 bp residue)

If we assume our conclusion of these fixed differences is correct, and they are completely diagnostic of the two species, then we can resolve some possible confusion of species identities listed in Genbank For example, GenBank access no: AJ553910.1 lists C gigas as the species, but it has, according to our work, nucleotides exclusive for C angulata and this ‘error’ would lead to confused phylogenetic trees Moreover, six COX1 sequences under name of C angulata sampled from the south of China from GenBank obviously belong to C gigas as they carry all of the five exclusive nucleotides for C gigas (details in Supplementary file S3_Table 1)

In this study, we confirm with extensive data that C angulata and C gigas, though very closely genetic related, can be categorically distinguished using mtDNA COX1 sequences differences at particular nucleotide positions This is the first report on diagnostic nucleotides for identifying C angulata and C gigas based on analysis of large number of COX1 sequences

Is the cultured oyster in Vietnam C gigas?

In this study, we aligned COX1 sequences from 72 samples taken from three Vietnamese lines (Ria1, Namdinh and China) with published COX1 sequences of C angulata and C gigas as references, and found all five exclusive nucleotides for C angulata were present in all Vietnamese samples from the three lines We conclude that cultured oysters from northern Vietnam are C angulata, not C gigas Hybrids between the two taxa can be fertile (Huvet et al.,

2002), and even a natural hybrid stock was established in some places e.g Tokyo Bay (Iwasaki et al., 2009) If there had been mixing (hybridization) of the two species in Vietnam, then we would conclude all females in the hybrid crosses must have been C angulata This seems unlikely on the basis of probability Also, the DNA microsatellite data, discussed following, did not indicate a substantial excess of heterozygotes perhaps expected if the animals sampled from Vietnam were hybrids These findings are agreement with the conclusions of a previous study that an oyster population in Taiwan, the origin of the Vietnamese RIA1 line, is pure C angulata population (Hsiao et al., 2009) Putative C angulata was found in many places in the South of China Sea (Liu et al., 2011, Wang et al., 2010, Xia et al., 2009) including Fujian and Guangdong provinces from where the China line in this study originated

There is accumulating biogeographic evidence that C gigas may favour sea water temperatures cooler than those in the north of Vietnam (which are 20-35 o C), and so North Vietnam is not suitable for C gigas to normally grow Indeed attempts to grow out C gigas in Vietnam during 2005/06 from brood-stocks imported from Australia failed (RIA1 annual report,

2012) On the other hand, C angulata favour hotter water temperature than C gigas (Orton and Awati, 1926) and so may have better tolerance to tropical water temperature found in northern Vietnam and the south of China

Of the three northern Vietnamese lines, the origin of two is known (China from China, and Ria1 from Taiwan), but the origin of the Namdinh line is unknown A previous study using COX1 sequences by Binh and Quyen (2013) could not resolve the species identity of the Namdinh line, however, this matter was resolved in this study Even so, the present study sheds no light on the precise origin of Namdinh line – it is possible it originated from either China or Taiwan

Three new C angulata haplotypes were found in Namdinh (Haplotype i1, i2) and China line (Haplotype i2, i3), but the frequencies of these haplotypes are low in these lines, 8.3%; 12.5% and 1.4% for C angulata haplotype i1, i2 and i3, respectively Finding a new haplotype is relatively rare in C angulata lines Huvet et al (2000b) used COX1 sequences (710 nucleotides) to analyse nine populations (five populations from Europe and three populations from Asia), but no new haplotypes previously reported were observed even when a large geographic zone was considered Boudry et al (1998) reported that haplotype A was the most common one (76%) in

C angulata samples studied, however in this study, only 25% (18 out of 72 samples of C angulata) carry C angulata haplotype A

What oyster species is being cultured in Nhatrang?

It appears that samples from Nhatrang belong to clades other than C angulata or C gigas All COX1 sequences from Nhatrang aligned with those reported either for Kumamoto oyster C sikamea or the Indian oyster, C madrasensis, and were obviously different from C angulata or

C gigas This finding corresponds with the report of Dang and Quyen (2003) who concluded, using COX1 sequences, that C sikamea was present in samples taken from Nhatrang (Khanh Hoa province) Hedgecock et al (1999) reported wild C sikamea occurred on the eastern and northern shores of Japan and it is assumed C sikamea, originates from southern Japan and is now present in many places in the world including China, Korea, USA and some European countries (Yu and Li, 2012) C sikamea looks like C gigas; these species cannot be distinguished

Conclusion

a) Five diagnostic nucleotides in mtDNA COX1 sequences have been confirmed in this study and can used to distinguish Crassostrea angulata from Crassostrea gigas b) The Vietnam aquaculture “Pacific oyster” in northern Vietnam is C angulata, not C gigas whereas local hatchery oyster samples in Nhatrang, Vietnam are a mixture of two oyster species: Crassostrea sikamea or Crassostrea madrasensis c) All three hatchery oyster lines (Ria1, Namdinh and China) still maintain adequate genetic diversity in short term for a start of a breeding program.

Acknowledgements

We gratefully acknowledge financial support from the University of the Sunshine Coast (USC) and Australian Centre for International Agricultural Research (ACIAR) through John Allwright Fellowship for financial support to this study and to V-V I We highly appreciate the technical assistance from USC staff and students, especially David Bright, Rob Lamont, Ido Bar and Dan Powell in instruction of lab techniques and genetic soft-wares.

Supplementary data

Supplementary file S3 Crassostrea angulata allelic and haplotypic data.

References

AMOS, W & HARWOOD, J 1998 Factors affecting levels of genetic diversity in natural populations

APPLEYARD, S A & WARD, R D 2006 Genetic diversity and effective population size in mass selection lines of Pacific oyster (Crassostrea gigas) Aquaculture, 254, 148-159

ASTANEI, I., GOSLING, E., WILSON, J & POWELL, E 2005 Genetic variability and phylogeography of the invasive zebra mussel, Dreissena polymorpha (Pallas) Molecular Ecology, 14, 1655-1666

BANKS, M., MCGOLDRICK, D., BORGESON, W & HEDGECOCK, D 1994 Gametic incompatibility and genetic divergence of Pacific and Kumamoto oysters, Crassostrea gigas and C sikamea Marine

BATISTA, F., LEITÃO, A., HUVET, A., LAPÈGUE, S., HEURTEBISE, S & BOUDRY, P The taxonomic status and origin of the Portuguese oyster Crassostrea angulata (Lamark, 1819) International Oyster Symposium, 2005 2005

BIJMA, P., VAN ARENDONK, J A & WOOLLIAMS, J A 2000 A general procedure for predicting rates of inbreeding in populations undergoing mass selection Genetics, 154, 1865-1877

BINH, T D & QUYEN, H V D 2013 5 th International Oyster Symposium (IOS5) Ho Chi Minh city, Vietnam:

BOUDRY, P 2008 Review on breeding and reproduction of Europhean aquaculture species Aqua

BOUDRY, P., COLLET, B., CORNETTE, F., HERVOUET, V & BONHOMME, F 2002 High variance in reproductive success of the Pacific oyster (Crassostrea gigas, Thunberg) revealed by microsatellite-based parentage analysis of multifactorial crosses Aquaculture, 204, 283-296 BOUDRY, P., HEURTEBISE, S., COLLET, B., CORNETTE, F & GÉRARD, A 1998 Differentiation between populations of the Portuguese Oyster, Crassostrea angulata and the Pacific Oyster, Crassostrea gigas, revealed by mtDNA RFLP analysis Journal of Experimental Marine Biology and Ecology,

BOUDRY, P., HEURTEBISE, S & LAPÈGUE, S 2003 Mitochondrial and nuclear DNA sequence variation of presumed Crassostrea gigas and Crassostrea angulata specimens: a new oyster species in Hong Kong? Aquaculture, 228, 15-25

BUROKER, N., HERSHBERGER, W & CHEW, K 1979 Population genetics of the family Ostreidae II

Interspecific studies of the genera Crassostrea and Saccostrea Marine Biology, 54, 171-184 CARLSSON, J., MORRISON, C L & REECE, K S 2006 Wild and Aquaculture Populations of the Eastern

Oyster Compared Using Microsatellites Journal of Heredity, 97, 595-598

DANG, B T & QUYEN, H V D 2003 5th International Oyster Symposium (1085), World Oyster Society

2013 Ho Chi Minh city, Dec 11-13, 2013

DAVID, D.-C & SAVINI, D 2011 Molecular approaches to bivalve population studies: a review Genetics and Molecular Biology, 12, 1-14

DE DONATO, M., MANRIQUE, R., RAMIREZ, R., MAYER, L & HOWELL, C 2005 Mass selection and inbreeding effects on a cultivated strain of Penaeus (Litopenaeus) vannamei in Venezuela

EVANS, B., BARTLETT, J., SWEIJD, N., COOK, P & ELLIOTT, N 2004a Loss of genetic variation at microsatellite loci in hatchery produced abalone in Australia, Haliotis rubra and South Africa,

EVANS, F., MATSON, S., BRAKE, J & LANGDON, C 2004b The effects of inbreeding on performance traits of adult Pacific oysters, Crassostrea gigas Aquaculture, 230, 89-98

FOIGHIL, D., GAFFNEY, P., WILBUR, A & HILBISH, T 1998 Mitochondrial cytochrome oxidase I gene sequences support an Asian origin for the Portuguese Oyster, Crassostrea angulata Marine Biology, 131, 497-503

FOLMER, O., BLACK, W H., HOEH, W., LUTZ, R & VRIJENHOEK, R 1994 DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates Molecular

FRANKEL, O H & SOULE, M E 1981 Conservation and Evolution Cambridge University Press, Cambridge,

FRANKHAM, R 1996 Relationship of genetic variation to population size in wildlife Conservation Biology,

GENE CODES CORPORATION, I 2011 Sequencher version 5.0 sequence analysis software Gene Codes

GOYARD, E., ARNAUD, S., VONAU, V., BISHOFF, V., MOUCHEL, O., PHAM, D., WYBAN, J & BOUDRY, P

2003 Residual genetic variability in domesticated populations of the Pacific blue shrimp (Litopenaeus stylirostris) of New Caledonia, French Polynesia and Hawaii and some management recommendations Aquatic Living Resources, 16, 501-508

GRANLEESE, T., CLARK, S A., SWAN, A A & VAN DER WERF, J H 2015 Increased genetic gains in sheep, beef and dairy breeding programs from using female reproductive technologies combined with optimal contribution selection and genomic breeding values Genetics Selection Evolution, 47, 1- HEASMAN, M P., GOARD, L., DIEMAR, J & R.B, C 2000 Improved early survivals of mollusks: Sydney Rock 13

Oyster, NSW Fisheries Report, NSW, Australia: Port Stephens Fisheries Center, 2003

HEDGECOCK, D., LI, G., BANKS, M & KAIN, Z 1999 Occurrence of the Kumamoto oyster Crassostrea sikamea in the Ariake Sea, Japan Marine Biology, 133, 65-68

HERSHBERGER, W K., PERDUE, J A & BEATTIE, J H 1984 Genetic selection and systematic breeding in

HSIAO, S.-T., LIANG, H.-Y., LIU, D.-C & SU, W.-C 2009 Using DNA barcodes to assess oyster biodiversity of Taiwan The 3rd International Oyster symposium (IOS3): Recent advances, potentials, challenges and problems in oyster industry 2-4 November 2009, Taipei, Taiwan: The World Oyster

HUVET, A., BOUDRY, P., OHRESSER, M., DELSERT, C & BONHOMME, F 2000a Variable microsatellites in the Pacific Oyster Crassostrea gigas and other cupped oyster species Animal Genetics, 31, 71-72 HUVET, A., FABIOUX, C., MCCOMBIE, H., LAPEGUE, S & BOUDRY, P 2004a Natural hybridization between genetically differentiated populations of Crassostrea gigas and C angulata highlighted by sequence variation in flanking regions of a microsatellite locus Marine Ecology Progress Series,

HUVET, A., GÉRARD, A., LEDU, C., PHÉLIPOT, P., HEURTEBISE, S & BOUDRY, P 2002 Is fertility of hybrids enough to conclude that the two oysters Crassostrea gigas and Crassostrea angulata are the same species? Aquatic Living Resources, 15, 45-52

HUVET, A., HERPIN, A., DÉGREMONT, L., LABREUCHE, Y., SAMAIN, J.-F & CUNNINGHAM, C 2004b The identification of genes from the oyster Crassostrea gigas that are differentially expressed in progeny exhibiting opposed susceptibility to summer mortality Gene, 343, 211-220

HUVET, A., LAPEGUE, S., MAGOULAS, A & BOUDRY, P 2000b Mitochondrial and nuclear DNA phylogeography of Crassostrea angulata, the Portuguese Oyster endangered in Europe

IN, V.-V., O'CONNOR, W., DOVE, M & KNIBB, W 2016 Can genetic diversity be maintained across multiple mass selection lines of Sydney rock oyster, Saccostrea glomerata despite loss within each?

IWASAKI, T., TANAKA, T., LIDZUKA, Y., HSIAO, S.-T & ARANISHI, F A biallelic microsatellite DNA analysis for hybrid oysters between Crassostrea gigas and C angulata The 3rd International Oyster symposium (IOS3): Recent advances, potentials, challenges and problems in oyster industry

2009, Taipei, Taiwan, 2-4 November: The World Oyster Society

JONES, O R & WANG, J 2010 COLONY: a program for parentage and sibship inference from multilocus genotype data Molecular Ecology Resources, 10, 551-555

KLINBUNGA, S., KHAMNAMTONG, N., TASSANAKAJON, A., PUANGLARP, N., JARAYABHAND, P &

YOOSUKH, W 2003 Molecular Genetic Identification Tools for Three Commercially Cultured Oysters (Crassostrea belcheri, Crassostrea iredalei, and Saccostrea cucullata) in Thailand Marine

KNIBB, W., WHATMORE, P., LAMONT, R., QUINN, J., POWELL, D., ELIZUR, A., ANDERSON, T., REMILTON,

C & NGUYEN, N H 2014 Can genetic diversity be maintained in long term mass selected populations without pedigree information? A case study using banana shrimp Fenneropenaeus merguiensis Aquaculture, 428–429, 71-78

LALLIAS, D., BOUDRY, P., LAPEGUE, S., KING, J W & BEAUMONT, A R 2010 Strategies for the retention of high genetic variability in European flat oyster (Ostrea edulis) restoration programmes

LAM, K & MORTON, B 2003 Mitochondrial DNA and morphological identification of a new species of

Crassostrea (Bivalvia: Ostreidae) cultured for centuries in the Pearl River Delta, Hong Kong, China Aquaculture, 228, 1-13

LAM, K & MORTON, B 2006 Morphological and mitochondrial-DNA analysis of the Indo-West Pacific rock oysters (Ostreidae: Saccostrea species) Journal of molluscan studies, 72, 235-245

LAPEGUE, S., BATISTA, F., HEURTEBISE, S., YU, Z & BOUDRY, P 2004 Evidence for the presence of the

Portuguese oyster, Crassostrea angulata in northern China Journal of Shellfish Research, 23, 759-

LARSEN, J B., FRISCHER, M E., RASMUSSEN, L J & HANSEN, B W 2005 Single-step nested multiplex PCR to differentiate between various bivalve larvae Marine Biology, 146, 1119-1129

LAUNEY, S., BARRE, M., GERARD, A & NACIRI-GRAVEN, Y 2001 Population bottleneck and effective size in Bonamia ostreae-resistant populations of Ostrea edulis as inferred by microsatellite markers

LI, G., HUBERT, S., BUCKLIN, K., RIBES, V & HEDGECOCK, D 2003 Characterization of 79 microsatellite

DNA markers in the Pacific oyster Crassostrea gigas Molecular Ecology Notes, 3, 228-232

LIND, C E., EVANS, B S., KNAUER, J., TAYLOR, J J & JERRY, D R 2009 Decreased genetic diversity and a reduced effective population size in cultured silver-lipped pearl oysters, Pinctada maxima

LIU, J., LI, Q., KONG, L., YU, H & ZHENG, X 2011 Identifying the true oysters (Bivalvia: Ostreidae) with mitochondrial phylogeny and distance based DNA barcoding Molecular Ecology Resources

LOPERA-BARRERO, N M., POVH, J A., RIBEIRO, R P., GOMES, P C., JACOMETO, C B & DA SILVA LOPES,

T 2008 Comparison of DNA extraction protocols of fish fin and larvae samples: modified salt (NaCl) extraction Ciencia e Investigación Agraria, 35, 65-74

MARSHALL, T., SLATE, J., KRUUK, L & PEMBERTON, J 1998 Statistical confidence for likelihood-based paternity inference in natural populations Molecular ecology, 7, 639-655

MASAOKA, T & KOBAYASHI, T 2005 Species identification of Pinctada imbricata using intergenic spacer of nuclear ribosomal RNA genes and mitochondrial 16S ribosomal RNA gene regions Fisheries

MENZEL, R 1974 Portuguese and Japanese oysters are the same species Journal of the Fisheries Board of Canada, 31, 453-456

MORTON, B., LAM, K & SLACK-SMITH, S 2003 First report of the European flat oyster Ostrea edulis, identified genetically, from Oyster Harbour, Albany, south-western Western Australia Molluscan

MUSTAFFA, S., L., A M., ARIFFIN, A H., M.N., D & NOR, S A M 2010 Cytochrome oxidase I (COI) gene highlights taxonomic ambiguities of Malaysian Crassostrea (Sacco, 1897) oyster species The 7th

IMT-GT UNINET and the 3rd Joint International PSU-UNS Conferences, Proceedings, 7-8 October

2010, Prince of Songkla University, Hat Yai, Songkhla, Thailand / Prince of Songkla University

NGUYEN, D 2009 Assessing genetic diversity in cultured aquatic species: the Sydney Rock Oyster,

Saccostrea glomerata stock improvement program as a model Masters by research thesis,

NICHOLAS, K B., NICHOLAS, H & DEERFIELD, D 1997 GeneDoc: analysis and visualization of genetic variation Embnew news, 4

O'CONNOR, W A., DOVE, M C., FINN, B & O’CONNOR, S J 2008 Manual for hatchery production of sydney rock oysters, Saccostrea glomerata IDP NSW NSW Department of Primary Industries Port

Stephens Fisheries Center Taylors Beach NSW 2316, Australia http://www.dpi.nsw.gov.au/research/areas/aquaculture/outputs/2008/oconnor3

O'FOIGHIL, D., GAFFNEY, P., WILBUR, A & HILBISH, T 1998 Mitochondrial cytochrome oxidase I gene sequences support an Asian origin for the Portuguese oyster Crassostrea angulata Marine Biology, 131, 497-503

ORTON, J H & AWATI, P R 1926 Modification by Habitat in the Portuguese Oyster Ostrea (Gryphcea) angulata Journal of the Marine Biological Association of the United Kingdom (New Series), 14,

PEAKALL, R & SMOUSE, P E 2012 GenAlEx 6.5: genetic analysis in Excel Population genetic software for teaching and research-an update Bioinformatics, 28, 2537-2539

PONZONI, R W., KHAW, H L., NGUYEN, N H & HAMZAH, A 2010 Inbreeding and effective population size in the Malaysian nucleus of the GIFT strain of Nile tilapia, Oreochromis niloticus Aquaculture,

PRITCHARD, J K., WEN, W & FALUSH, D 2003 Documentation for STRUCTURE software: version 2

RADULOVICI, A E., ARCHAMBAULT, P & DUFRESNE, F 2010 DNA Barcodes for Marine Biodiversity:

REECE, K., RIBEIRO, W., GAFFNEY, P., CARNEGIE, R & ALLEN, S 2004 Microsatellite marker development and analysis in the eastern oyster (Crassostrea virginica): confirmation of null alleles and non- Mendelian segregation ratios Journal of Heredity, 95, 346-352

REECE, K S., CORDES, J F., STUBBS, J B., HUDSON, K L & FRANCIS, E A 2008 Molecular phylogenies help resolve taxonomic confusion with Asian Crassostrea oyster species Marine Biology, 153, 709-721

RIA1 ANNUAL REPORT 2012 Bivalve production in Vietnam Annual report

SANIL, N., SUJA, G., LIJO, J & VIJAYAN, K 2012 First report of Perkinsus beihaiensis in Crassostrea madrasensis from the Indian subcontinent Diseases of Aquatic Organisms (DAO), 98, 209-220

SCHMIDT, R 2011 Evaluation of GeneMapper® ID-X and GeneMarker® HID for use at the NYC OCME SOLETCHNIK, P., HUVET, A., LE MOINE, O., RAZET, D., GEAIRON, P., FAURY, N., GOULLETQUER, P &

BOUDRY, P 2002 A comparative field study of growth, survival and reproduction of Crassostrea gigas, C angulata and their hybrids Aquatic Living Resources, 15, 243-250

STEPIEN, C., MORTON, B., DABROWSKA, K., GUARNERA, R., RADJA, T & RADJA, B 2001 Genetic diversity and evolutionary relationships of the troglodytic ‘living fossil’Congeria kusceri (Bivalvia:

SUZANA, M., LUTFI, A M., HADI, A A., DEVAKIE, M & AZIZAH, M S 2011 Genetic variation in Malaysian oysters: taxonomic ambiguities and evidence of biological invasion Biological Invasions, 13, 1893-

TAMURA, K., STECHER, G., PETERSON, D., FILIPSKI, A & KUMAR, S 2013 MEGA6: molecular evolutionary genetics analysis version 6.0 Molecular biology and evolution, 30, 2725-2729

TARIS, N., ERNANDE, B., MCCOMBIE, H & BOUDRY, P 2006 Phenotypic and genetic consequences of size selection at the larval stage in the Pacific oyster, Crassostrea gigas Journal of Experimental Marine Biology and Ecology, 333, 147 - 158

VAN OOSTERHOUT, C., HUTCHINSON, W F., WILLS, D P & SHIPLEY, P 2004 MICRO-CHECKER: software for identifying and correcting genotyping errors in microsatellite data Molecular Ecology Notes,

WANG, H., QIAN, L., LIU, X., ZHANG, G & GUO, X 2010 Classification of a common cupped oyster from southern China Journal of Shellfish Research, 29, 857-866

WANG, Y & GUO, X 2008 ITS length polymorphism in oysters and its use in species identification Journal of Shellfish Research, 27, 489-493

XIA, J., YU, Z & KONG, X 2009 Identification of seven Crassostrea oysters from the South China Sea using

PCR–RFLP analysis Journal of Molluscan Studies, 75, 139-146

YU, H & LI, Q 2012 Complete mitochondrial DNA sequence of Crassostrea nippona & comparative and phylogenomic studies on seven commercial Crassostrea species Molecular Biology Reports, 39, 999-1009

YU, Z & GUO, X 2004 Genetic analysis of selected strains of eastern oyster (Crassostrea virginica Gmelin) using AFLP and microsatellite markers Marine biotechnology, 6, 575-586

Are strain genetic effect and heterosis expression altered with culture system

Introduction

World mollusc production including clams, oysters, mussels, scallops and abalone, has reached 15.2 million tons, and accounts for approximately 23% of total aquaculture production (Astorga, 2014) The Portuguese oyster, Crassostrea angulata, is a species that originated from the Northwest Pacific region and became a key aquaculture species in Europe from the late 19 th century through to the early 1970’s due to its fast growth and high survival rate (Batista et al.,

2005) The Portuguese oyster was eventually replaced by the Pacific oyster (Crassostrea gigas) due to disease sensitivity and mass mortality in commercial production systems (Boudry, 2008) Recently, the oyster under cultivation in Vietnam, thought to be C gigas, was identified as C angulata (In et al., 2016b) One of the advantages of C angulata is that they better suited to the warm water temperatures (20-35 o C) encountered in northern Vietnam compared to Pacific oysters (Ria1, unpublished data) This might be the reason for the existence and dominance of an industry based on culture of C angulata in Asia, especially China, Taiwan and Vietnam

In northern Vietnam, C angulata production began in 2006-2007 and by 2010-11 it had reached 7,000 tons, worth A$7.0-9.8 (FIS/2010/100, 2013) With this rapid development, however, problems arose, in particular, farmers complained of poor quality stock, thought to be due to inbreeding To develop and sustain the oyster sector, a selective breeding program was commenced to deliver genetic improvement for specific traits as well as manage inbreeding In

2013, three different broodstock populations that had originated from Vietnam, Taiwan and China were collected and held at RIA 1 National Marine Broodstock Centre (CatbaMBC), Catba Island, Vietnam, for evaluation

Strain evaluations can involve testing the performance differences among strains (termed

“strain additive genetic effects”) and heterosis expression (from crosses between lines), and whether these measures may be altered with culture system and environment (Pongthana et al., 2010; Thoa et al., 2014) Crossbreeding among genetically different populations may yield an

111 increase in allelic heterozygosity and decrease the proportion of homozygous genotypes for deleterious recessive and lethal alleles leading to an increase in heterosis or hybrid vigour (Whitlock et al., 2000) This method (crossbreeding) can enhance the productivity of farmed stocks and is considered an alternative genetic approach to selective breeding for specific traits Crossing among lines (often in the form of complete diallel crosses) can also broaden the genetic variability in newly created synthetic (mixed) populations and so foster long-term response to selection Such experiments have been reported for Placopecten magellanicus (Wang and Côté,

2012), Meretrix meretrix (Dai et al., 2014), Haliotis discus hannai (Deng et al., 2010), Salmo salar (Gjedrem et al., 1991), Genetically Improved Farmed Tilapia (GIFT) Oreochromis niloticus

(Bentsen et al., 1998, Eknath et al., 2007, Nguyen et al., 2010) and Oreochromis niloticus (Thoa et al., 2014)

In this study, we carried out full 3 x 3 diallel crosses to assess genetic components of three oyster strains in Vietnam The diallel cross approach was applied to: 1) permit statistical separation of the additive genetic component from non-genetic (heterosis or hybrid vigour) effects, 2) permit the estimation of reciprocal (maternal) effects and 3) foster the creation of a new synthetic mixed line if needed By testing progeny of crosses in different culture systems, it was possible to estimate the effect of genotype by environment interaction Accounting for non- additive genetic effects in genetic evaluation models is important in order to obtain unbiased estimates of the additive genetic values of strains as well as individuals within strains and thus to maximise response to selection (Falconer and Mackay, 1996)

In the present study, we estimated strain additive genetic value, heterotic and maternal effects on harvest body weight of three C angulata populations and their crossbreeding families that were grown in two different sites including Catba Islands in Haiphong and Vandon Islands in Quangninh province in northern Vietnam At each site, oysters were cultured following two methods including “single” and “cultch” cultures The principal aim of this study was to establish a base population for a future genetic breeding program for C angulata in Vietnam.

Materials and methods

The three populations of C angulata included in this study were: 1) a strain from Ria1 hatcheries on Catba Islands (Haiphong) known to have been derived from imported Taiwanese

112 stocks (Ria1 line); 2) local stocks from Namdinh province that were being cultured at Vandon Islands, Quangninh in Vietnam (Namdinh line) and 3) a third strain was imported from China at the beginning of 2013 (China line) The three populations were held at the Northern National Broodstock Centre for Mariculture, RIA1 in Catba Islands, Haiphong City (CatbaMBC, Figure 5.1) under common rearing conditions In October 2014 (10 months), when oyster stocks reached maturity at an average weight of 50.5 ± 4.5g for males and females, a complete diallel cross involving the three strains (50 males to 100 females per population) was conducted Parents that had high quality eggs (fully developed, ovular/ovular/around eggs) or sperm (dense motile suspensions) from each population were used for the diallel cross combination

Figure 5.1 Grow-out sites of C angulata

All broodstock strains (Ria1, Namdinh, China) were kept at CatbaNBC before the grow- out trials reported here The three populations were maintained under the same temperature, salinity and food supply conditions Oyster families produced from the diallel crosses were reared in the hatchery facility of the CatbaNBC for approximately two months until they had reached a size of approximately 2.5 – 3.0 mm before they were transferred to sea cage sites for grow-out at Vandon Islands in Quangninh province and Catba Islands in Haiphong city, Vietnam of Catba NBC (Figure 5.1)

One day prior to spawning, all three strains (Ria1, China and Namdinh) cultured in sea cages at Catba Islands in Haiphong City were transported to the CatbaNBC facility where they were conditioned for one day in 8m 3 concrete tanks at ambient temperature (25-27 o C) and a salinity of 30‰ Oyster families were generated by a strip-spawning For a given cross, 1 L seminal fluid (mixture of sperm and seawater of 30‰ salinity) from one male were divided into two equal parts in 2.5 L plastic buckets (buckets 1 and 2) at room temperature (26-27 o C) All eggs from the first female were fertilized with sperm from bucket 1 to make the first full-sib family and all eggs from the second female were fertilized with sperm from bucket 2 to make the second full-sib family The first and the second full-sibs belong to one half-sib family Consequently, this mating formula will make two full-sib and one half-sib families The mix of sperm and eggs was gently mixed, and then filtered through a 200 àm mesh screen to remove debris before transfer to a single, labelled 500 L larval rearing tank Fertilized eggs were incubated in these tanks overnight at room temperature (25-27 o C) and D-veliger larvae were observed after 20h After that, all D-veliger larvae from each tank (one tank per family) were siphoned in to a 40 àm mesh bag and then transferred to clean tanks containing 200L of seawater The larvae were reared at approximately 50 larvae/L, and fed a mixture of Nannochloropsis oculata, Isochrysis galbana and

Chlorella sp, twice daily Initially, N oculata was the main feed for larvae in the first week of rearing at approximate 30,000 cells algae/mL Then, after one week an equal ratio of these three algal species were supplied to the rearing tanks

Rearing water was exchanged at a rate of 20-30% per day in each tank and added water was dosed with algae Every three days, larvae were collected on mesh screens and transferred to a new tank containing 200L of clean seawater When pediveligers were observed (approximately day 20), the larvae of each family were divided equally One group was used for the production of “cultch” spat by introducing monofilament strings with 70-80 dead oyster shells on each string for larval settlement The remaining larvae were used to produce “single” (cultch-free spat) by holding larvae in a downwelling system After settlement, oyster spat of both types were retained in the hatchery for 40 days until they had reached 2.5 - 3.0 mm in length before they were transferred to grow-out sites

5.2.4 Grow-out testing environments and harvest measurement

Two types of seed were tested (Single seed and Cultch set) Single seed are individual oysters not attached to a substrate, while cultch set are oysters that have been allowed to settle on the surface of an oyster shell These two settlement methods require different cultivation practices

For single seed: the individual oysters were grown in baskets at 5 spat/cm 2 The bags were then hung in seawater under a floating bamboo raft (3x3x3m) In order to prevent oysters escaping or dropping out, a net was used to cover the baskets Periodically, the net was changed to bigger net size in order to permit greater water flow to increase food availability

For cultch culture: Spat settled on shells of dead oysters were hung at intervals on a nylon string One string held 8-10 shells, with each shell having approximately 20 spat settled on it Strings were suspended from floating bamboo rafts (3x3x3m)

Both “single” and “cultch” spat were farmed on Vandon Island, whereas only “cultch” were held at Catba Island Water temperature at both locations ranged from 15-35 o C, while salinity was between 26 and 35 ‰ The baskets used for culturing oyster families were cleaned and dried for 1 h, monthly to remove debris and reduce fouling Every second month, 30 oysters from each family, both in “cultch” and “single” culture types, were randomly sampled to measure the body weight

Following a grow-out period of approximately 270 days, all oysters reached marketable size At this time thirty samples were randomly taken from each family to measure shell length and body weight

A total of 7,269 oysters were recorded at harvest from a 3 x 3 complete diallel cross involving three C angulata populations (China, Namdinh and RIA1) at CatbaNBC, Research Institute for Aquaculture No.1 In this analysis, pure strains (China, Namdinh and RIA1) was calculated separately for each location or farming types A general linear model procedure in SAS (SAS, 1997) was used for initial analyses to test the statistical significance of systematic effects on body weight The systematic fixed effects were cross, culture environment, type of culture and their two-way interaction Stocking weight did not differ from crosses and therefore, the partial R 2 value stayed unchanged when it was fitted in the model As a result, final models did

116 not include stocking weight to analyse harvest body weight The mathematical expression of the final model is written as follows (Myer et al., 2010): yilklmnp = à + Gi + Ej + Tp + (G x E)ij + (G x T)ip + (E x T)jp + eilklmnp (1)

Where yilklmnp is the measurement of body weight, à is constant, Gi is the fixed effect of cross (cross combination, i = 6, three pure strains and three crosses), Ej is the test environments (j = 2 associated with two culture environment locations including Vandon Islands in Quangninh and Catba Islands in Haiphong in Vietnam), T is the type of culture (p = 2, “single” or “cultch”), G x E is the interaction between cross and environment, G x T is the cross by type of culture interaction, E x T is the environment by culture type interaction, eilklmnp is the random residual term

To estimate strain additive genetics and heterosis, the effect of genotypes included in model 1 was partitioned into direct additive, non-additive and reciprocal (maternal) effects The model 1 become (Myer et al., 2010): yijkp = à + Ej + Tp + (E x T)jp + ∑α iai + ∑α ij hij + ∑βiri + eilkp (2) where E, T are the fixed effects of environment, type of culture respectively, α i is the proportion of genes attributed by the nth individual derived from the i th strain (α i = 0.0, 0.5 or 1.0 and ∑α i = 1.0); ai is the additive genetic effect of animal genes derived from the ith strain; αij is the coefficient of the total hybrid vigour effect for the cross combination between the ith and jth strains (αij = 0.0 or 1.0; i ≠ j; ij ≠ ji and ∑αij = 1.0); hij is the total of hybrid vigour effect for the cross combination between the i th and j th strains (i ≠ j; ij ≠ ji); the general maternal effect coefficient is βi for i th strain (βi = 0 for purebreds and -0.5 for male strain and 0.5 for female strain, for the crossbreds and ∑βi = 1.0); ri is the general maternal effect of the i th strain and eilkp is the vector with random residual effects for the animal

The partial F test was used to evaluate the significance of the additive, heterosis, and maternal effects by removing each term from the full model at a time (Myers et al., 2010) The additive genetic, heterotic and maternal effects were estimated as regression coefficients with one degree of freedom The additive genetic effects were limited to ∑α i = 0 The number of sires and dams was restricted per cross, therefore, the general maternal effect coefficients set in the current study suppose that the additive genetic effects of a given strain are the same regardless of the sex of parents (Gjerde et al., 2002) The maternal effects included in the model attempted to improve the accuracy of the additive genetic and hybrid vigour predictions Total heterosis of a two-strain cross was calculated by formula: hij = ĥ + hi + hj + sij, where ĥ is the average heterosis

Results

Descriptive statistics for harvest body weight of C angulata in two locations and two culture systems are shown in Table 5.1 The average body weight at harvest did not differ significantly between experimental locations for “cultch” culture system (50.7 g in Catba in comparison with 46.9 g in Vandon, P>0.05) By contrast, the harvest body weight (46.9g) in the

“cultch” culture was two times greater than “single” culture system (22.5g) at Vandon The coefficients of variation in body weight were slightly greater in Vandon than in Catba

Table 5.1 Descriptive statistics for harvest total weight of a 3 x 3 diallel cross population of C angulata oyster in two different culture environments and methods (g)

Culture environment Type of culture N Mean SD CV%

N = actual number of oyster at harvest that is different from at the initial stocking time, SD = standard deviation and

5.3.2 Effects of culture system (“single” vs “cultch”) and environment (Catba vs Vandon)

The statistical analysis of systematic fixed effects on harvest total weight is shown in Table 5.2 The type of culture and its interaction with cross-combination (nine crosses in total) had

118 statistically significant effects on the trait studied (P < 0.0001) In contrast, effects of culture locations (environments) including Vandon and Catba, as well as cross-combination between the three strains were not statistically significant for the harvest total weight

Table 5.2 Statistical significance of fixed effects on harvest total weight of oyster

Data Effects Degree of freedom

The body weight at harvest of the three pure strains of C angulata and their reciprocal crosses at two different environments and two types of culture are shown in Figures 5.2 and 5.3 Across the two environments, there are no significant differences (P>0.05) in harvest body weight among the three pure strains and their crosses (Figure 5.2), whereas the body weight at harvest in “cultch” culture is significantly higher than in “single” system (Figure 5.3, P0.05) This could be explained by hatchery practice when farmers often used broodstock from various sources as they did not have enough mature animals from their own stocks for breeding As a result, this practice may have introduced new alleles and haplotypes from different groups functioning as sublines described in Knibb et al (2014) and In et al (2016) and so helped recovered the genetic loss.

Evaluate strain genetic effect and heterosis

Findings in Chapter 3 allowed a breeding trial using the three lines of C angulata (Ria1, Namdinh and China) In this study, we investigated if strain performances (breed additive genetic effect) and heterosis of harvest body weight varied between culture environments (Catba vs

Namdinh) and types of culture (single vs cultch) Data from this study indicated that the rankings of strain/breed effect regarding additive genetic performance did not differ significantly between systems studied (P>0.05), perhaps because of the similar condition in these two locations (Catba vs Vandon) or types of culture (“single” vs “cultch”) Regardless of culture systems and locations, the China strain showed highest additive genetic performance while Ria1 was poorest The phenotypic observations here, arguably, are consistent with our findings in Chapter 3 that the line from China had the highest number of allele and haplotypes, whereas Ria1 stock was the lowest Despite no re-ranking effect observed between grow-out locations or culture conditions, there were considerable differences in additive genetic values for harvested body traits among the three lines regarding rearing location and type of culture This could be due to different genetic backgrounds and different histories of selection Estimated heterosis for harvested body weight and growth was low and not different among the rearing locations (Catba vs Vandon) and between the types of culture (“single“ vs “cultch”)

The reciprocal (maternal) effects not only relate to growth but also to fitness traits such as survival as commonly observed during the larval development in shellfish species (Deng et al.,

2010) In the present study, the maternal effects differed among the three strains (Namdinh, Ria1, and China) and between testing environments (two different locations) and culture systems (“single“ vs “cultch”) To exploit these results, and notwithstanding the issue of forming an outbred synthetic line, perhaps the strain from Namdinh can be used as a maternal strain, whereas the one from China should be included as a paternal line.

References

APPLEYARD, S A & WARD, R D 2006 Genetic diversity and effective population size in mass selection lines of Pacific oyster (Crassostrea gigas) Aquaculture, 254, 148-159

BERNAY, B., BAUDY-FLOC'H, M., ZANUTTINI, B., ZATYLNY, C., POUVREAU, S & HENRY, J 2006

Ovarian and sperm regulatory peptides regulate ovulation in the oyster Crassostrea gigas

BIGOT, L., ZATYLNY-GAUDIN, C., RODET, F., BERNAY, B., BOUDRY, P & FAVREL, P 2012

Characterization of GnRH-related peptides from the Pacific oyster Crassostrea gigas Peptides,

BOUDRY, P., HEURTEBISE, S., COLLET, B., CORNETTE, F & GÉRARD, A 1998 Differentiation between populations of the Portuguese Oyster, Crassostrea angulata and the Pacific Oyster,

Crassostrea gigas, revealed by mtDNA RFLP analysis Journal of Experimental Marine Biology and Ecology, 226 , 279-291

CARLSSON, J., MORRISON, C L & REECE, K S 2006 Wild and Aquaculture Populations of the Eastern

Oyster Compared Using Microsatellites Journal of Heredity, 97, 595-598

CHANSELA, P., SAITONGDEE, P., STEWART, P., SOONKLANG, N., STEWART, M., SUPHAMUNGMEE, W.,

POOMTONG, T & SOBHON, P 2008 Existence of APGWamide in the testis and its induction of spermiation in Haliotis asinina Linnaeus Aquaculture, 279 , 142-149

CROPPER, E C., MILLER, M W., TENENBAUM, R., KOLKS, M., KUPFERMANN, I & WEISS, K R 1988

Structure and action of buccalin: a modulatory neuropeptide localized to an identified small cardioactive peptide-containing cholinergic motor neuron of Aplysia californica Proceedings of the National Academy of Sciences, 85 , 6177-6181

DANG, B T & QUYEN, H V D 2003 5th International Oyster Symposium (1085), World Oyster Society

2013 Ho Chi Minh city, Dec 11-13, 2013

DENG, Y., LIU, X., ZHANG, G & WU, F 2010 Heterosis and combining ability: a diallel cross of three geographically isolated populations of Pacific abalone Haliotis discus hannai Ino Chinese Journal of Oceanology and Limnology, 28 , 1195-1199

EVANS, B., BARTLETT, J., SWEIJD, N., COOK, P & ELLIOTT, N 2004 Loss of genetic variation at microsatellite loci in hatchery produced abalone in Australia, Haliotis rubra and South Africa,

HEDGECOCK, D., LI, G., BANKS, M & KAIN, Z 1999 Occurrence of the Kumamoto oyster Crassostrea sikamea in the Ariake Sea, Japan Marine Biology, 133, 65-68

HSIAO, S.-T., LIANG, H.-Y., LIU, D.-C & SU, W.-C 2009 Using DNA barcodes to assess oyster biodiversity of Taiwan The 3rd International Oyster symposium (IOS3): Recent advances, potentials, challenges and problems in oyster industry 2-4 November 2009, Taipei, Taiwan:

HUVET, A., BOUDRY, P., OHRESSER, M., DELSERT, C & BONHOMME, F 2000 Variable microsatellites in the Pacific Oyster Crassostrea gigas and other cupped oyster species Animal Genetics, 31, 71-72

IN, V.-V., O'CONNOR, W., DOVE, M & KNIBB, W 2016 Can genetic diversity be maintained across multiple mass selection lines of Sydney rock oyster, Saccostrea glomerata despite loss within each? Aquaculture, 454 , 210-216

KATAOKA, H., TOSCHI, A., LI, J P., CARNEY, R L., SCHOOLEY, D A & KRAMER, S J 1989 Identification of an allatotropin from adult Manduca sexta Science, 243 , 1481-1483

KNIBB, W., WHATMORE, P., LAMONT, R., QUINN, J., POWELL, D., ELIZUR, A., ANDERSON, T.,

REMILTON, C & NGUYEN, N H 2014 Can genetic diversity be maintained in long term mass selected populations without pedigree information? — A case study using banana shrimp

LIND, C E., EVANS, B S., KNAUER, J., TAYLOR, J J & JERRY, D R 2009 Decreased genetic diversity and a reduced effective population size in cultured silver-lipped pearl oysters, Pinctada maxima Aquaculture, 286 , 12-19

LIU, J., LI, Q., KONG, L., YU, H & ZHENG, X 2011 Identifying the true oysters (Bivalvia: Ostreidae) with mitochondrial phylogeny and distance based DNA barcoding Molecular Ecology Resources MILLER, M W., BEUSHAUSEN, S., CROPPER, E C., EISINGER, K., STAMM, S., VILIM, F S., VITEK, A.,

ZAJC, A., KUPFERMANN, I & BROSIUS, J 1993 The buccalin-related neuropeptides: isolation and characterization of an Aplysia cDNA clone encoding a family of peptide cotransmitters J Neurosci, 13 , 3346-57

NAGLE, G T., PAINTER, S., BLANKENSHIP, J & KUROSKY, A 1988 Proteolytic processing of egg- laying hormone-related precursors in Aplysia Identification of peptide regions critical for biological activity Journal of Biological Chemistry, 263, 9223-9237

NAMBU, J R & SCHELLER, R H 1986 Egg-laying hormone genes of Aplysia: evolution of the ELH gene family The Journal of Neuroscience, 6 , 2026-2036

O'FOIGHIL, D., GAFFNEY, P., WILBUR, A & HILBISH, T 1998 Mitochondrial cytochrome oxidase I gene sequences support an Asian origin for the Portuguese oyster Crassostrea angulata

REECE, K S., CORDES, J F., STUBBS, J B., HUDSON, K L & FRANCIS, E A 2008 Molecular phylogenies help resolve taxonomic confusion with Asian Crassostrea oyster species Marine Biology, 153 ,