Bonamia -free ﬂ at oyster ( Ostrea edulis L.) seed for restoration projects: non-destructive screening of broodstock, hatchery production and test for Bonamia -tolerance

– Native ( ﬂ at) oyster ( Ostrea edulis ) beds, once a major component of the North Sea, largely disappeared from the region in the late 19th century. Flat oyster restoration is taking place at a number of locations in the North Sea. When ﬂ at oyster beds are restored in areas where O. edulis is functionally extinct it is advised to treat these as disease-free areas. Adult oysters were collected in the Dutch Delta area which is infected with the pathogenic parasite Bonamia ostreae . The aim of this research was to obtain Bonamia -free seed from parents collected in a Bonamia -infected area. In addition, the oysters were analysed to identify candidate genomic regions related to bonamiosis tolerance and exposed to Bonamia in the ﬁ eld to assess survival compared to a naïve control group. With the aid of a non-destructive screening method, Bonamia - free broodstock were selected. These oysters produced Bonamia -free larvae and seed. For comparison, broodstock oysters were collected in the Dutch Wadden Sea, an area free of Bonamia . These oysters also produced Bonamia -free larvae and seed. To study if the Delta area oysters had developed a degree of resistance to the disease, while the naïve Wadden Sea oysters had not, seed of both groups was challenged in Lake Grevelingen where Bonamia occurs. Survival of the pre-selected Bonamia -free oysters was signi ﬁ cantly higher than the naïve group. Samples of seed were analysed for association of candidate genetic markers related to bonamiosis tolerance. A higher percentage of individuals with tolerance-associated marker genotypes was found in the screened group compared to the naïve one. However, mortality of the naïve group could not be related to Bonamia presence. Further challenge tests are needed before ﬁ rm conclusions regarding the genetic markers can be made. The results show that hatchery production of Bonamia -free and potentially Bonamia -tolerant ﬂ at oysters is possible.


Introduction
The restoration of habitat-forming species is being applied throughout the world's oceans to halt the loss of biodiversity, combat climate change and help sustain the provision of ecosystem services such as nutrient cycling or eutrophication control (Saunders et al., 2020). The flat oyster Ostrea edulis L. was a common species in the North Sea until about a century ago (Gercken and Schmidt, 2014;Houziaux et al., 2008;Olsen, 1883). However, intensive flat oyster fishery at the end of the 19th century caused the oyster population to decline rapidly (Gercken and Schmidt, 2014;Houziaux, 2008). Currently, the species is listed as 'threatened' or 'declining' by the where the disease has been present for a long period, tolerance seems to have been developed (Culloty et al., 2004;Lynch et al., 2014). Other countries, such as Norway, are Bonamia free (Mortensen et al., 2016). When flat oyster beds are restored in areas where O. edulis is functionally extinct it is advised to treat these as Bonamia-free areas (Pogoda et al., 2019).
In Europe, there are 12 hatcheries that produce flat oyster seed for aquaculture purposes (Colsoul et al., 2021;Kamermans et al., 2020). These hatcheries are either located in Bonamia-infected areas or Bonamia-free areas (Sas et al., 2020). Seed from Bonamia infected areas cannot be used in restoration project because of the risk of spreading the disease (Pogoda et al., 2019). Seed from Bonamia-free hatcheries has the disadvantage that the seed is vulnerable to the disease (Culloty et al., 2004;Lynch et al., 2014). If, for some reason, Bonamia will reach the restoration site, the oysters may suffer from high mortality. Therefore, the use of disease-free seed produced from disease-tolerant parents would solve this problem if tolerance is inheritable. This offers opportunities to produce oysters that are not infected with Bonamia, despite the parents having been exposed to Bonamia, potentially creating offspring with some tolerance to the disease.
In the Netherlands, B. ostreae was first observed in the Oosterschelde in 1980 and in Lake Grevelingen in 1988 (Engelsma et al., 2010). This implies that the oysters present in these waters have been exposed to the disease for a long period. Culloty et al. (2004) compared the performance of oysters that had been selectively bred for resistance to B. ostreae (Rossmore, Cork harbour, Ireland), and oysters from two areas where Bonamia had been present for a long time (Lake Grevelingen, the Netherlands and Brittany, France), with oysters from four naïve populations where Bonamia had not been detected (Lough Foyle, Ireland; Tralee, Ireland; Lough Kishorn, Scotland; Mull, Scotland). Strains from these seven populations were translocated to Cork Harbour (Ireland), Lake Grevelingen (the Netherlands) and Brittany (France). The field trials indicated that Rossmore and Lake Grevelingen oysters showed lower mortality compared to other populations. Culloty et al. (2004) conclude that previous exposure in these populations had conferred some reduced susceptibility to the parasite compared to naïve populations. In a follow up study, seed was produced in the hatchery of Roem van Yerseke with broodstock from long-term exposed populations in Lake Grevelingen and the Oosterschelde and a naïve population in Limfjord in Denmark. Seed of all three groups was reared for one year in Lake Grevelingen. Survival was best in seed from Lake Grevelingen (OYSTERECOVER, 2013). It was concluded that the oyster stock from Lake Grevelingen should be considered as a candidate stock for starting a breeding plan in the Netherlands. Although this stock had the highest overall prevalence of infection, it also had the greatest growth and survival rate indicating that it may have formed some local tolerance to the disease. In addition, Vera et al. (2019) identified SNP markers associated with Bonamia tolerance in flat oyster using distant naïve and tolerant European populations and these markers were recently validated in a narrower area of the North Sea (Sambade et al., 2022). Since most of these markers mapped at a single genomic region, they could constitute a useful tool to facilitate obtaining Bonamiatolerant stocks.
The aim of this research was to obtain Bonamia-free seed from parents collected in a Bonamia-infected area. In addition, the oysters were analysed to identify candidate genomic regions related to bonamiosis tolerance and exposed to Bonamia in the field to assess survival compared to a naïve control group.
2 Materials and methods 2.1 Non-destructive screening of broodstock oysters Broodstock oysters of around 5 years old were collected from the Delta area by the hatcheries of Roem van Yerseke and Stichting Zeeschelp. Prior to screening they were taken out of their storage tanks and left dry at 8-14°C for 24 h. The next day, the oysters were anaesthetised by placing them in 5 L containers using 50 g L À1 magnesium chloride dissolved in a mixture of fresh water (3 L) and seawater PSU 31 (2 L) at 15-20°C to maintain salinity (Suquet et al., 2010). A small amount of algae (Isochrysis galbana) was added regularly to the water to stimulate filtration by the oysters and the water was aerated. After approximately 3 h in the anaesthetic solution, the status of the oysters was assessed. Oysters were considered anaesthetised when their valves remained open after exerting pressure. A piece of gill of approximately 4 mm 2 was cut with scissors and placed in a vial containing seawater. The vial and the oyster were given the same code number. Following the sampling, oysters were placed in separate containers with clean seawater until their valves closed. Then the valves were kept closed with a rubber band and a piece of paper with the number was slipped under the rubber band. The oysters were placed at 7°C without water for 3-4 days, until the result of the Bonamia analysis (described below) became available. Based on the Bonamia analysis the oysters were then separated into two groups: Bonamia free and Bonamia infected. Bonamia free oysters were transferred to the hatcheries where larvae and seed were produced following Helm et al. (2004). The screening was carried out with five broodstock batches in the period 2018-2021. The two hatcheries subsequently provided larvae and seed produced from these screened non-infected broodstock for Bonamia analysis and field challenges.

Test of presence of Bonamia in gills vs rest of the body
For routine monitoring of B. ostreae in the flat oyster population in the Netherlands a pool of oyster tissue was collected from gill, mantle and digestive diverticulum (up to a total of ∼25 mg tissue per individual oyster), below referred to as "standard method". In order to investigate whether the nondestructive method of sampling (gill tissue of approximately 4 mm 2 ) is providing a representative status of the presence of Bonamia in the oyster, both methods were compared with the real-time PCR as described below using 190 oysters collected in 2018, 2019 and 2021 during the routine monitoring for shellfish diseases from Lake Grevelingen.

Survival test when exposed to Bonamia
Flat oyster seed produced in 2018 by the Royal Netherlands Institute of Sea Research (NIOZ) and by Roem van Yerseke (RvY) were placed in the field in February 2020. The NIOZ seed was produced with broodstock from the Wadden Sea (Jacobs et al., 2020). This area is not known to be infected with Bonamia and the seed is therefore considered to be naïve. The RvY seed was produced at the hatchery of Roem van Yerseke with selected Bonamia-free broodstock from the Oosterschelde. At the start of the experiment, each group consisted of around 2000 individuals.While waiting for the possibility to transfer to an oyster culture plot Lake Grevelingen (from February 2020 until August 2020) the two groups were kept in one oyster basket (BST Oyster Supplies) each. These two baskets were attached to a FLUPSY (floating upwelling system) of Roem van Yerseke in the harbour of Yerseke (Oosterschelde). From August 2020 onwards the baskets were placed in two frames that kept four oyster baskets each suspended in a vertical way above a bottom culture plot for flat oysters in Lake Grevelingen. Each basket consisted of two individually numbered compartments and per group 1 to 7 baskets were used (see Supplementary  Table 2). Stocking density was 74-114 oysters per compartment. In February 2021 a second group of oysters from the same 2018 NIOZ batch was transferred to the baskets in Lake Grevelingen. Stocking density was 82-148 oysters per compartment. Every 6 weeks the live oysters were counted and a sample of 60 individuals per group was placed at À20°C for Bonamia and tolerance-associated markers analysis. Before storage a picture of the oysters was taken with a ruler in view. Shell width was determined of 30 individuals per group with the programme ImageJ. Survival was based on the counts and expressed as percentage of the previous count. Differences between groups were tested with t-test per date. The challenge test lasted until August 2021. Both the Oosterschelde and Lake Grevelingen are areas that show Bonamia infection (Van Banning, 1987;Haenen and Engelsma, 2020).

Bonamia analysis
Samples of oyster gill tissue, larvae and seed were homogenised with a TeSeE PRECESS 24 homogenizer (Bio-Rad) in tissue disruption tubes equipped with ceramic beads. DNA was extracted from the homogenised material and water samples using the MagNA Pure LC JE379 (Roche Diagnostics, Almere, Netherlands) with the MPLC DNA isolation kit 2 according to the protocol of the manufacturer or the NucliSens EasyMAG (bioMerieux Benelux bv, Zaltbommel, Netherlands) according to Specific protocol B of the manufacturer. The DNA was eluted in 50 ml and stored at À20°C until further processing. The detection of genetic material of Bonamia by real time PCR was carried out based on the real time PCR described by Marty et al. (2006). The realtime PCR amplification reactions were performed using an ABI 7500 Fast Applied Biosystems Thermocycler (Applied Biosystems). All reactions were carried out in a total volume of 20 ml in a 96-wells plates (Bioplastics) consisting of 10 ml of TaqMan PCR Master Mix (Applied Biosystems), 0.4 mM of the forward-primer (5'-CCCGGCTTCTTAGAGGGACTA-3'; Applied Biosystems), 0.4 mM of the reverse-primer (5'-ACCTGTTATTGCCCCAATCTTC-3'; Applied Biosystems), 0.08 mM of the Minor Groove Binding probe (5'-6-FAM-CTGTGTCTCCAGCAGAT-NFQ-3'; Applied Biosystems), 0.25 ml Uracil-DNA Glycolase (UDG 5 U/ml; New Engeland Biolabs), 5 ml template DNA and distilled water to make up a final volume of 20 ml. Each real-time PCR run included nontemplate controls (with 5 ml distilled water as template) as well as negative and positive controls (extracted DNA of bivalve individuals known to be infected or free of B. ostreae). The following thermal conditions were used: 37°C for 10 min; 95°C for 10 min and 40 cycles of 95°for 3 s, 60°C for 30 s. At the end of each cycle the fluorescence signal was collected during the annealing step (60°C). As control of the integrity of the DNA extraction a parallel reaction was carried out for each sample using a Eukaryotic 18S rRNA Endogenous Control mix (Applied Biosystems). The threshold cycle (Ct) value, an increase of fluorescence above the background, was calculated by the Sequence Detection Software version 1.5 of the 7500 System Software (Applied Biosystems).

Genotyping of SNP markers associated with Bonamia tolerance
The 22 SNPs previously reported to be consistently associated with divergent selection for Bonamia tolerance (Vera et al., 2019) were evaluated to be genotyped in a single multiplex using a MassARRAY platform (Sequenom, San Diego, CA). Briefly, the technique consists of a two-step reaction: The first involves the PCR amplification of an amplicon which includes the selected SNP (i.e. locus-specific PCR reaction) and the second a single-base extension using mass-modified dideoxynucleotide terminators of an oligonucleotide primer that anneals immediately upstream of the polymorphic site (i.e. SNP) of interest (Oeth et al., 2009;Ellis and Ong, 2017). To design the set of primers for MassARRAY genotyping, the oligo-nucleotides of the oyster chip, where the SNPs are located (Gutierrez et al., 2017), were mapped in the flat oyster genome recently assembled and annotated NCBI Bioproject: PRJNA772111). Then, regions of ±100 bp flanking the selected SNPs were retrieved and used for primer design (Supplementary Table 1). After in silico primer design evaluation, 18 SNPs could be included in a single multiplex reaction harnessing the distinct mass of extended primers adjacent to the SNP position of different mass (Supplementary Table 1). MALDI-TOF mass spectrometry analysis in an Autoflex spectrometer was used for allele scoring. This genotyping was conducted at the UCIM-University of Valencia Genomics Platform to validate 2b-RAD genotyping.

Analysis of genetic structure in the sample collection
The number of different genetic clusters (K value) in the sampling collection, composed by 381 individuals from seed produced by RvY (192) and the seed produced by NIOZ (189) collected during the challenge test in April 2021, was evaluated with STRUCTURE v 2.3.4 (Pritchard et al., 2000), a program that analyses linkage disequilibrium and Hardy-Weinberg equilibrium following a Bayesian clustering approach. For each of the possible number of clusters tested (from 1 to 10), the R package ParallelStructure v 1.0 (Besnier and Glover, 2013) was used with a burn-in of 100,000 iterations and 200,000 Markov Chain Monte-Carlo steps, as well as ten independent replicate runs to increase accuracy. To identify the most likely number of clusters (K), two different K estimators were used: deltaK (Evanno et al., 2005) and Mean LnP(K) (Pritchard et al., 2000). StructureSelector web-based software (Li and Liu, 2018) was used to obtain K estimators and CLUMPAK graphical outputs (Kopelman et al., 2015). This analysis was performed with the 16 most consistently related to bonamiosis tolerance SNPs.

Screening method
In order to test whether the non-destructive method of sampling for Bonamia is comparable in sensitivity to the standard analysis (including a piece of gill, mantle and digestive tissue), as described above, 190 Lake Grevelingen oysters were tested with both methods. Bonamia genetic material was detected in 40 individuals out of the 190 oyster for each method: the standard sampling method and the nondestructive sampling method (Tab. 1). However, both methods had positive samples which were not detected by the other method. It should be noted that in all positive oysters only low amounts of Bonamia genetic material was detected (real time PCR Ct values ≥ 32), which might have caused the false negatives. It can be concluded that the non-destructive sampling is comparable for detecting a Bonamia infection.
The non-destructive screening of broodstock oysters identified 8-32% of the tested oysters as Bonamia infected (Tab. 2). The oysters that tested Bonamia negative were used to produce larvae and seed. All larval and seed batches were Bonamia free (Tab. 2).

Survival and growth in a Bonamia-infected area
For the first 8-12 weeks in both 2020 and 2021 survival was close to 100% for both origins ( Fig. 1 and Supplementary Table 2). In 2020, survival of the Bonamia-naïve offspring (NIOZ) rapidly decreased between 19 May and 7 August, with only 25% survival. After that sampling date survival increased again. The survival of the pre-screened Bonamia-free broodstock (RvY) oysters also decreased, but much more gradually. For these oysters survival at the end of the challenge experiment was 73% (Fig. 1). In 2021, survival at the end of the experiment was higher than 2020 with 92% for the RvY oysters and 54% for the NIOZ oysters. In 2020, the highest mortality occurred in July, while June showed the highest mortality in 2021. The offspring of the pre-screened Bonamiafree broodstock (RvY) showed significantly higher survival than the offspring of Bonamia-naïve broodstock (NIOZ) in the second part of the season in both years (Fig. 1). The initial size . From February until August 2020 the two groups were kept in the harbour of Yerseke (Oosterschelde) in one basket each. From August onwards 1 to 7 baskets per group with two compartments each were placed in Lake Grevelingen. * = significant difference between the two groups, P < 0.05. of the NIOZ oysters was significantly smaller than the RvY oysters (Fig. 2). However, from August 2020 onwards the average sizes were comparable, except for April 2021, when RvY oysters were larger.

SNP markers associated with Bonamia tolerance
The genotypes of the 381 individuals studied for the 18 SNP markers are shown in Supplementary Table 3. Nine individuals (five from RvY and four from NIOZ) showed missing genotyping call in ≥ 6 SNPs (corresponding with ≥ 1/3 of molecular markers analysed) and they were discarded. Therefore, the number of individuals used for population genetics analyses was finally 372.
Following the deltaK method proposed by Evanno et al. (2005) and the Mean LnP(K) described by Pritchard et al. (2000), the best K value in the samples analysed was 2. The Bonamia-free sample, coming from the seed produced with pre-screened broodstock from a long-term affected area by RvY (187 individuals: 48 "orange"/139 "blue"), showed much higher proportion of individuals belonging to the "orange" cluster than the sample coming from seed produced with broodstock from a naïve area by NIOZ (185 individuals, 7 "orange"/178 "blue") and the difference was highly significant (exact test P-value < 0.00001) (Fig. 3).

Bonamia presence in challenged oyster
Samples were analysed for presence of Bonamia at three sampling dates over a period of 8 months. All samples were negative for Bonamia (Tab. 3). This is surprising as it was expected that the NIOZ group, coming from a naïve area and . From February until August 2020 the two groups were kept in the harbour of Yerseke (Oosterschelde). From August onwards they were placed in Lake Grevelingen. * = significant difference between the two groups, P < 0.05. with a lower percentage of genetic variants associated with Bonamia tolerance, should be expected to become infected with Bonamia. As a result, a relationship between individual presence, or absence, of markers and level of Bonamia infection could not be established.

Discussion
The results showed that the non-destructive screening method provides information on the status of the oyster with regard to Bonamia infection. In addition, Bonamia negative broodstock always produced Bonamia negative larvae and seed. De Melo et al. (2021) produced disease free seed of Pacific oysters (Crassostrea gigas) through after spawn testing of broodstock. Unlike C. gigas, O. edulis cannot be strip spawned. Therefore, a non-destructive method was developed. This provides opportunities for Bonamia-free hatchery production with broodstock from a Bonamia-infected area. A prerequisite is that production takes place in a Bonamia-free zone of the hatchery, as Flannery et al. (2016) suggested that larvae may be able to acquire the pathogen from the water column during filter feeding. Water treatment in the hatcheries involved filtration (5 mm) and UV.
The challenge experiment with pre-screened Bonamia free and Bonamia naïve flat oysters showed that the survival of the naïve group was less than the Bonamia-free group. In addition, the naïve group showed a significantly lower proportion of the "orange" cluster associated with Bonamia tolerance in previous studies (Vera et al., 2019;Sambade et al., 2022). However, none of the oysters became infected with the parasite during exposure in Lake Grevelingen according to the PCR test performed. There are some hypotheses that could explain this inconsistency. Firstly, the parasite was not present in the area. This seems unlikely since the area is long-term affected (Fig. 4). In 2020, no sampling was performed due to COVID-19 restrictions, but in 2021 Bonamia was detected in the oysters during the regular sampling program indicating that the parasite is still present in the area. Secondly, the test sensitivity could be too low. However, the PCR methodology itself is very sensitive, and the controls used in the DNA extraction and PCR test did not show any failures or reduced sensitivity. Thus it would be expected that if Bonamia was present it should have been detected, also from weakly infected oysters. Thirdly, the oysters did not contract the disease. It is largely unknown when B. ostreae disperses to infect other oysters. The naïve oysters were introduced in Lake Grevelingen in two batches and contraction of the disease perhaps took place in the period that naïve oysters were not present (7 December 2020 to 5 February 2021). Furthermore, it is unknownhow long it takes for the oysters to become infected and how, once infected, the proliferation of the parasite progress. From cohabitation experiments in which naive oysters are exposed to (heavily) infected oysters can be deduced that it takes several months before the parasite can be detected (Culloty et al., 2004;Lynch et al., 2005;Lallias et al., 2008;Ronza et al., 2018). The naïve oysters were deployed in Lake Grevelingen for four months in 2020 (August-December) and in 2021 for 6 months (February-August). This time may not have been long enough to get infected and/or development of the infection to a detectable level. Normally spawning stress causes more mortality in infected oysters (Engelsma et al., 2010), probably leading to increase Bonamia-free oysters as a result of dying. The question is if this is coinciding with the (yet unknown) time that Bonamia spreads. As shown in Figure 4 the percentage of infected B. ostreae infected flat oysters in 2021 was 26%. This was however the overall results from Lake Grevelingen. But this spread is not homogeneous, between the sites the distribution is more patchy. This might have Fig. 3. STRUCTURE plot using the 18 SNP markers associated with resilience to bonamiosis in the sampled individuals from seed produced with pre-screened broodstock from a long-term affected area by RvY and the seed produced with broodstock from a naïve area by NIOZ collected during the challenge test in April. Each vertical bar represents one individual, and the colour proportion for each bar/individual represents the posterior probability of its assignment to the different clusters (K) inferred by the program. Individuals pertaining to these two clusters, "orange" and "blue", showed most of them a full "orange" or "blue" constitution, and further, they appeared intermingled in both samples at very different frequencies. contributed to a lack of infection in the challenge population. Also keeping the oysters in suspended culture above the bottom may have reduced the likelihood of getting infected. Our conclusion is that the much higher mortality of the naïve oysters compared to the prescreened oysters was most likely caused by other factors such as a potential difference in condition at the start of the experiment.
The analysis of the genetic markers is a follow-up of the research of Sambade et al. (2022). In that study, the same groups of hatchery-produced seed from screened Bonamiafree broodstock and seed from Bonamia-naïve broodstock (from the Dutch Wadden Sea) were analysed for the potential association of the consistent outlier SNP loci reported by Vera et al. (2019) with bonamiosis resistance. In addition, the screened Bonamia-free broodstock that produced the seed was also analysed. The pre-selected Bonamia-free oysters and seed showed a higher percentage of individuals with the tolerance genetic constitution ("orange"), while the naïve oyster seed showed a lower percentage. The present study also detected a significant lower percentage of "orange" individuals in the naïve group as compared to the offspring from screened Bonamia-free broodstock. The higher percentage of "orange" genotypes associated with Bonamia tolerance of pre-screened Bonamia-free seed is corroborated by a higher survival of that group. These results give additional support to previous findings on the association of a specific genomic region with Bonamia tolerance (Sambade et al., 2022), but at the same time they cannot discard that other genomic regions could be involved on tolerance since association was detected at the population level. Unfortunately, none of the oysters tested positive for Bonamia. Thus, the desired quantitative information on parasite load could not be correlated with genotyping information in oysters from the field challenge, information which is crucial to establish the association at individual level. Further challenge tests are needed before firm conclusions regarding the genetic markers can be made.
If it is assumed that higher survival of pre-screened Bonamia-free seed is related to the higher percentage of the "orange" constitution related to bonamiosis tolerance, the question remains whether this is resistance or tolerance. Holbrook et al. (2021) consider disease-resistance to be reduced susceptibility to infection by the parasite, or active suppression of the parasites ability to multiply and proliferate. They define disease-tolerance as the retention of fitness and an ability to neutralise the virulence of the parasite. Tolerance also includes the possibility of maintaining health despite being infected. Disease-resilience is identified as the ability to recover from illness. Råberg et al. (2008) states that resistance should reduce the prevalence of the parasite in the host population (causing selection in both the parasite and the host), while tolerance should have a neutral or positive effect on parasite prevalence. Since Bonamia prevalence in Lake Grevelingen is not declining over time, the higher survival and presence of candidate genetic variants found in the preselected Bonamia-free oysters compared to the naïve oyster indicate tolerance and not resistance.
Producing Bonamia-free oysters with broodstock originating from a Bonamia-infected area provides disease-free oysters that may also have developed tolerance to the disease. This is very useful for restoration projects since transfer of diseases is not wanted, but protection against disease is desired, in case it does show up in a newly established bed.
Acknowledgements. This research was funded by the Ministry of Agriculture, Nature and Food safety (LNV) (BO43-18 and KB34), Programma Rijke Waddenze (PRW) and De Rijke Noordzee (DRN). We thank Betty van Gelderen and Michal Voorbergen-Laarman (WBVR) for analysing the Bonamia samples, Emiel Brummelhuis, Ad van Gool, Jack Perdon, Wouter Suykerbuyk and Annamara de Wolf (WMR), Harry Heidekamp (LNV) and the crew of the vessel Regulus for help with the field sampling and oyster farmer Arjen van Vliet for making the poles of his culture plot available for the oyster baskets. The Ministry of Agriculture, Nature and Food safety, PRW and DRN were not involved in the study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the article for publication.

Supplementary Material
Supplementary Table1. Information about SNPs markers used in the present study. Name of the SNP marker within Sequenom multiplex (SNP marker), name of the marker used by Vera et al. (2019) (Vera et al. IDs), genomic region where marker is located (Sequences, SNP marker and its variants are shown between brackets), left primer, right primer and extension primer for the Massarray Sequenom multiplex are shown. The SNPs removed are highlighted in red. Table 2. Number of oysters counted in BST baskets attached to a FLUPSY or to a frame above a bottom plot (numbered compartments) and the number of oysters collected for analysis in 2020 (A) and 2021 (B).

Supplementary
Supplementary Table 3. Genotyping from the 381 initial individuals. "-" indicates no genotyping call. The criteria for eliminating individuals was 6 or more SNPs markers with no genotyping call. Individuals removed from genetic analyses are shown in red.