Mass Production of American Oyster Spat (Crassostrea virginica) and Development of Improved Broodstock through use of High-density Pools for Production of Optimum Quality Larvae

Final Report
New Brunswick Shellfish Growers Association (NBSGA)
AIMAP 2012-G05

Executive Summary

The proposed oyster spat production project consists in mass production of larvae produced in hatcheries in high-density pools developed in New Zealand and production of spat from improved broodstock from the only American oyster breeding program in Canada, namely, the program at the Coastal Zones Research Institute (CZRI). This project will enable the industry to benefit from improved oyster seed.

Production of oyster larvae using high-density pools began at CZRI in January 2013. The project will continue until late May and produce larvae from our broodstock under the selective breeding program. We are currently in Phase 2, Section B Adjustments and Adaptation, which was to have ended in December 2012. As a reminder, here are the main sources of delay since the start of the project:

• The visit to New Zealand planned for April 2012 actually took place in June;
• The pools were delivered on November 6 but had been expected during the summer of 2012;
• As a result, the system could not be installed and put into operation until December 2012, which is when Phase 2 was to have ended.

Despite the delays, preliminary results suggest the future is promising for New Brunswick's oyster growing industry and the American Oyster selective breeding program at CZRI. Based on observations to date, more than two million larvae can be maintained per high-density pool, and the system in place includes 60 pools. In the CZRI's hatchery alone, there is the potential for more than 120 million eyed larvae per year (assuming no accidental mass mortality occurs). Following the project, it has become possible to substantially upgrade the oyster selective breeding program at the CZRI and transfer adapted larvae production techniques from New Zealand to industrial use. Similarly, spat optimized through genetic selection will benefit commercial hatcheries and growers in their efforts to deal with the increasing needs of the oyster growing industry. These processes will result in better quality and higher performance seed and thus promote better commercial productivity.

1. Introduction¹

Aquaculture production in New Brunswick generates significant economic activity in rural and coastal regions and ranks second in Canada after British Columbia. In New Brunswick, American oyster production totalled $9.6 million in 2009 in terms of farm gate value and processing level. The shellfish industry in New Brunswick is made up of 276 bottom culture leases and 233 suspension culture leases. There were 108 million oysters in production in 2009 (DAAF/2009 Aquaculture Sector Overview). 

For the past five years, the Coastal Zones Research Institute (CZRI) has been conducting an American oyster (Crassostrea virginica) selective breeding program. This selective breeding program has enabled better mastery of the techniques for producing families through interbreeding between individuals and the production of three first-generation F1 cohorts in 2005, 2007 and 2012. The results of these research efforts showed that some families have a higher growth and survival potential than naturally collected oysters. Certain families have very interesting characteristics such as one family with a growth rate 10% higher than naturally collected oysters and a low mortality rate under 5%. Oysters from a pool of broodstock have higher growth rates at 20.8% compared to natural collection, representing a reduction of nine months in a four-year growth cycle. All of these observations indicate that there is true potential for obtaining better performing animals, which constitutes the foundation of this CZRI research program (CZRI 2010). The same observations were made in the literature review by geneticist Dr. Christophe Herbinger. He says that "[Translation] The results presented in the program report are quite encouraging. Family production techniques through male-female interbreeding were developed. Several families were produced and cultured and there were interesting growth differences in the field. These are substantial advances because controlled production of shellfish families for selective breeding programs is still rare in the world. Such programs exist in only a few countries (USA, France, Australia, New Zealand) and are still in the early stages. Establishing a selective program in New Brunswick will be a great asset for shellfish industry development in that province and in Atlantic Canada" (Herbinger 2010).

To ensure production of the required families for a selective breeding program, oyster seed must be produced in a hatchery. The CZRI hatchery team encountered some challenges during this development period in the selective program. The fertilization rate was quite variable and, in most cases, low. One hundred (100) hybrids and five pools were needed to produce sixteen (16) families. Of this group, only a few proved worthwhile in terms of growth and survival improvement. All this can be accounted for by the quality of different gametes between adults, as well as gametic incompatibility. To be able to continue this selective breeding program and develop it more quickly, some aspects must be reviewed, including that regarding rearing families at a hatchery. Larval rearing pools must be changed in order to produce considerably more families simultaneously as is done in other selective breeding programs involving oysters, with the same physical space and more larvae per family. Pools of this type were developed in New Zealand to enable their own Pacific oyster breeding program to be carried out. This is a key factor for faster and more efficient development of an American oyster breeding program in New Brunswick in order to bring significant benefits to the industry more quickly. One suggestion mentioned in the breeding program review states that the current hatchery, which is equipped with a small number of large pools in a closed system, must be modified. This system is inappropriate for the objectives and the new system, as presented in the report, is based on a large number of small pools in an open system (Herbinger 2010).

These pools, called "Cawthron ultra density larval systems (CUDLS)" were developed at the Cawthron Institute in New Zealand, where the pools are used for producing families in their Pacific oyster and green mussel breeding programs. These pools are based on the open-system principle as opposed to the conventional method of batch culture. This new type of pool offers several advantages:

• an increase in the number of larvae per mL (100 to 1,000 instead of 2 to 10);
• faster larval growth and more uniform sizes compared with conventional systems;
• better settlement success.

2. Material and Methods²

2.1 Phase 1: Technological Transfer of New Zealand Pools

Objective: To receive training on the installation, operation and use of the CUDLS

Method

• Nick King of the Cawthron Institute conducted training in New Zealand on pool use and installation.
• Purchase of high-density pools.

2.2 Phase 2: Installation, Adjustments and Adaptation of CUDLS

Installation of system

To ensure the pools were operating well and to meet the criteria for American oyster breeding, a plan (described below) was drafted and followed. The pools were installed in existing facilities in the CZRI's oyster hatchery. The system formerly in use remained in place for comparison. Since the new system is an open circuit, a water tank is necessary to ensure a continuous water supply and at the desired temperature. An alarm and a generator were also added to the system to ensure uninterrupted water supply (breakdowns, power outages, etc.).

Adjustments and adaptation of the system

The CUDLS were delivered to the CZRI on November 6, 2012, instead of in summer 2012 as planned. As a result, the system was only put into operation in December 2012. Tests for adjustments and adaptation of the system began subsequently.

Before animals can be reared in the CUDLS, the pools must be conditioned with salt water alternating with dry periods for a minimum of two weeks. In addition, after installation, pools were tested to determine adjustments and ensure it would be possible to:

1. Maintain the correct temperature;
2. Maintain the required flow;
3. Distribute feed uniformly;
4. Maintain the water level in the pools.

Accordingly, temperature was noted in a traditional pool and a CUDLS (with a flow of 120 mL/min.) on an hourly basis for 24 hours to determine whether the system could maintain an adequate temperature. In the meantime, the ambient temperature of the hatchery room was noted regularly. Various water flow levels were also assessed (80, 120 and 160 mL/min.) in the pools of a single section and checked twice weekly for two weeks to ensure they were maintained. The uniformity of feed distribution was done using a video camera with the microalga Isochrysis galbana in one of the CUDLS. Finally, the water level in the CUDLS was observed throughout these preliminary tests to determine whether it remained stable.

This conditioning time and system adjustments were concurrent with the conditioning period (about 5 to 6 weeks) for broodstock intended for the subsequent steps. 

To adapt the operating factors for the new pools, American oyster larvae were produced in the hatchery using the conventional method. Broodstock were randomly selected from the same naturally collected group. Three mass spawnings were used to obtain the same batch of individuals for each replication over time. 

There were three replications comparing the performance of this new pool system to that of the traditional system. It was decided to test two different densities of embryos in the CUDLS immediately after fertilization, i.e. 500 and 1,000 embryos/mL, to determine if it was possible to avoid using traditional pools for the pre-larval period. In the traditional pools there were 10 embryos/mL, the usual level.

For each replication, the larvae were distributed in two 115-litre traditional pools (in a static system) and 18 CUDLS with different initial larval densities. Larval density in the traditional pools was monitored mid-cycle (about 10 days) and at the end of rearing, with no reduction to the larval density in the CUDLS. A total of 54 CUDLS, including larval density and water flow treatments (6X3X3 = 54 pools), and six conventional pools (2X3 = 6 pools) were necessary for this part of the experiment.

From day 2 of larval development, the number of larvae per pool was evaluated to determine growth, survival and the percentage of fertilization for each of the treatments. At the larval stage, 50 larvae per pool were measured (or fewer for pools with a very low number of larvae). Survival was evaluated by counting the number of living larvae per pool at each water change (3 times a week).

The setting stage will use eyed oyster larvae and Chinese hat collectors. The number of oysters transferred will be noted. Once the oysters are set, the setting percentage will be evaluated for each of the treatments.

2.3 Phase 3: Second-generation Spat Production

After mastering the operation of the new system and determining the parameters necessary for its proper operation, production of second-generation eyed larvae will begin by end of May 2013.

Since 2005, 2007 and 2012, first-generation families of the selective breeding program have been reared at sea. Oysters from the 2005 and 2007 cohorts are mature and could produce the second generation under the American oyster selective breeding program. The various hybrids will be determined with the assistance of geneticist Christophe Herbinger. The crossbreeding plan is hierarchized by the males, i.e. 1 male with 4-6 females resulting in full and half-brothers. The family production protocol appears below. The new system preserves more eggs per female, a distinct advantage in a selective breeding program.

F2 Family Production Protocol

• Selection parameter: growth and survival

Fertilization

• Crossbreeding plan hierarchized by males (1 ♂ with 4-6 ♀ =  full and half-brothers).
• The choice of hybrids will be determined by the geneticist.
• Fertilization will occur in glass beakers.
• Number of eggs and sperm used for fertilization:

  Families: 3,000,000 oocytes and 200 spermatozoa /oocyte (females with fewer than 3,650,000 eggs will not be used).

  Pool: 650,000 oocytes from each female and 200 spermatozoa from each male/oocyte.

• Gametes will be obtained by stripping.
• First a male will be found and stripped and the seed filtered over 60 µm mesh. Sperm motility will be checked. Minimal water will be used to ensure the sperm is as concentrated as possible. If the sperm are active, their number will be evaluated using a hematocytometer before finding a female. The seed will be placed in the refrigerator.
• Four females will be found and stripped.
• Eggs will be filtered in a 130 µm sieve to remove debris and then again in a 20 µm sieve. Eggs kept in the 20 µm filtering will be used.
• Then, a 20-minute waiting period will allow eggs to round out (they will be stirred every 10 minutes).
• The number of eggs will be evaluated using a particle counter.
• There will be three counts per sample (3 different pans per sample).
• Fifty eggs will be collected.
• Dividing the eggs: Family (3,000,000) in 4-litre beakers and Pool (650,000) in a 10-litre bucket.
• Fertilize only Family eggs and keep sperm in the refrigerator for the Pool at the end.Check again that sperm are active before using them.
• Wait 1.5 hours for fertilization.
• Always stir up the eggs every 10 minutes.
• When the first 4 females are fertilized, completely clean the workspace before seeking another male and 4 more females.
• When all Families have been fertilized, proceed with the Pool.

Rearing larvae

There will be no grading to prevent eliminating potentially higher performance animals. The densities and various criteria for raising larvae will have been determined in the system adjustment phase. At each filtering:

• Size of 50 larvae per family (photo) with each water change.
• Density per family (3 counts within a known volume) at each water change.
• Water temperature and salinity each time pools are filled.

Metamorphosis and spat

When the larvae are ready for setting, they will be placed in the other pools with the collectors. Each family will be set separately. The setting percentage will be evaluated for each family. 

Collection of genetic tissue for broodstock

Samples will be taken from each female and male involved in spawning for the production of the F2s.

1. If the bottles have previously been used, they must be rinsed with ethanol.
2. Ensure instruments are well cleaned between dissections.
3. Fill bottles to the shoulder with ethanol.
4. Take a piece of muscle (see figure showing internal organs of an oyster).
5. IMPORTANT: the size of the piece must not exceed ½ cm3 (maximum 1 cm3). This is equivalent to the size of a pencil eraser.
6. Each tissue sample is preserved in 20-mL glass bottles at room temperature in absolute ethanol.
7. If there is no ethanol in the bottles, the tissue samples can be preserved at - 80ºC.

3. Results and Discussion

3.1 Phase 1: Technological Transfer of New Zealand Pools

The mission was planned for April 2012 but ultimately took place in June 2012. Chantal Gionet and André Dumas represented the CZRI on a scientific mission to New Zealand's Cawthron Institute last June. The purpose of the visit was knowledge transfer of technology related to the use of high-density pools to produce mollusc larvae. The CZRI acquired 62 of these pools for use in its oyster selective breeding program. The Cawthron Institute team provided full training on installation and use of the pools. These scientists are leaders in the genetic improvement of molluscs. The visit provided an opportunity to learn more not only about molluscs but also about managing innovation as our representatives met personally with the Aquaculture Group Manager and the Chief Executive of this nearly 100-year-old research institute.

3.2 Phase 2: Installation, Adjustment and Adaptation of the CUDLS

In the initial testing to determine whether the system could maintain the temperature in the CUDLS, ambient temperature varied between 18.2°C and 26.4°C, which was insufficient to achieve a stable temperature in the pools (20.6°C to 22.5°C) even if the water tank temperature was stable at 19°C. The heating system was replaced and upgraded, which resulted in a more stable and more accurate ambient temperature of 20.8°C when adjusted to 20.0°C. With heating adjusted to 20.5°C and the water tank at 21.1°C, the average temperature in one of the CUDLS was 20.4°C ± 0.2°C. A few lower temperatures in the pools may have been caused by the opening of the laboratory door, which temporarily lowered the ambient temperature of the room. It is important then to keep the doors closed to prevent this type of variation during the production period.

Distribution of the alga Isochrysis galbana was filmed as planned, but the results have not yet been viewed and analyzed in detail. However, when the algae concentration in the water distribution pool is adjusted to the desired level and remains stable, water exiting the CUDLS (in the absence of larvae) with the same flow seems to have an algae concentration very similar to the water distribution pool. This is positive and indicates that the pools are each receiving nearly the same quantity of feed. Advance testing indicated that the pools could maintain a stable flow over a two-week period. However, adjusting for a rate of 80 mL/min. in the pools was more difficult than for rates of 120 mL/min. and 160 mL/min. given the very low flow and the limited accuracy with the water intake valve in the pool. Moreover, when the system was operating with larvae, the flow in the pools, adjusted to 80 ml/min., slowed almost to the point of stopping. As a result, the flow had to be continually readjusted on a regular basis, which was not the case for pools with flow rates of 120 or 160 mL/min. Consequently, it is not recommended to use a flow rate below 120 mL/min. per pool with the system in place.

Initially, bubbling in the pools was very low (about 1 bubble per second) to imitate procedures in New Zealand. However, sometimes bubbling stopped completed the day after the adjustment, which often cause the pool to overflow if the larvae and algae were lumping in the filter. So bubbling that was 10 times more powerful proved preferable--even essential--to reduce the risks of bubbling stopping and the filter getting clogged inside the pool. Despite the more powerful bubbling, the larvae continued to grow normally and no obvious mortality has been associated with it so far.

The filter in the CUDLS initially had 20 µm mesh to eliminate any chance of larvae escaping through the filter. However, a day or two after the larvae were placed in the pools, the filters of several pools clogged, which led to overflows. To counter this, 45 µm filters were installed. A slight loss of larvae occurred in the days following the transfer but, based on mafnification observations with a binocular loupe, it appears only very small or misshapen larvae passed through the 45 µm membrane. After 10 days of raising the larvae, it is recommended to install 60 µm filters; 45 µm filters become dirty more rapidly after about 10 days, particularly in high-density pools, which raises the risk of overflow.

On day 9, larval development seems more significant in the CUDLS than in the traditional pools. In fact, the average for larvae in CUDLS with a density of 1,000 larvae/mL and a flow of 120 mL/min. is 192 µm ± 17µm (n=10), while the average measurement for a traditional pool is 135 µm ± 10 µm (n=10). This trend remains to be confirmed in future production of larvae.

3.3 Phase 3: Second-generation Spat Production

This section of the protocol could not be evaluated in the planned period, but testing should be completed by May 2013.

4. Conclusion

As testing of larval rearing began late, only one replication has been undertaken until now and only up to day 15 of rearing. The setting stage is planned shortly for the first larvae. During rearing, the CUDLS was improved in terms of bubbling and filters in the pools, which reduced the risk of overflows. Testing continues and the larval density and water flow tests will be repeated twice more. Subsequently, production of spat from the second generation can be attempted under the best breeding conditions determined in the initial testing. The number of first-generation families will also be increased according to the geneticist's recommendations.